VOLUME VII SLICE V Cosway to Coucy (continued 4)

Until the success of the water-frame, cotton could not be spun economically of sufficient strength and fineness for warps, and the warps were therefore invariably made of either linen or wool. Some were manufactured locally, others were imported from Germany, Ireland and Scotland. The weaver prepared them for his loom by the system of peg-warping,19 but after the introduction of the warping-mill he received them as a rule all ready for insertion into the loom from the Manchester merchant or local fustian master.

“It did not pay the individual weaver to keep a warping-mill for occasional use only, and frequently the contracted space of his workroom precluded even the possibility of his doing so. The invention of the warping-mill necessitated specialism in warping, and it was essential that warping should be done to order, since at that time, the state of the industrial world being what it was, no person could ordinarily have been found to adventure capital in producing warps ready made in anticipation of demand for the great variety of fabrics which was even then produced. Moreover, had the weaver himself placed the orders for his warps, any occasional delay in the execution of his commissions might have stopped his work entirely until the warps were ready; for warps cannot be delivered partially, like weft, in quantities sufficient for each day’s work. To ensure continuous working in the industry, therefore, it was almost inevitable that the merchant should himself prepare the warps for such fabrics as he required, or possibly have them prepared. To the system of the merchant delegating the preparation of warps there was less objection than to the system of the weaver doing so, since the merchant, dealing in large quantities, was more likely to get pressing orders completed to time. Further, the merchant knew first what kind of warps would be needed. The first solution, however, that of the merchant undertaking the warping himself, was the surer, and there was no doubt as to its being the one destined for selection in a period when a tendency to centralize organization, responsibility and all that could be easily centralized, was steadily gaining in strength.”20

Guest says the system by which the weaver was supplied with warps and other material was substituted for the purchase of warps and cotton-wool by the weaver about 1740. No doubt the change was very gradual, especially as Aikin mentions the use of warping-mills in the 17th century. The weaver as a rule received his weft material in the form of cotton-wool and was required to arrange himself for its cleaning and spinning. According to Aikin,21 dealers tried the experiment of giving out weft instead of cotton-wool, but “the custom grew into disuse as there was no detecting the knavery of the spinners till a piece came in woven.” As it was impossible to unwrap the yarn and test it throughout its length, defects were hidden until it came to be used, and the complaints of weavers were not conclusive as to the inferiority of the yarn, since their own bad workmanship might have had something to do with its having proved unsatisfactory. It was therefore found best to saddle the weaver with full responsibility for both the spinning and weaving. Women and children cleaned, carded and spun the cotton-wool in their homes. The cotton had to be more thoroughly cleaned after its arrival in this country. The ordinary process of cleaning was known as “willowing,” because the cotton was beaten with willow switches after it had been laid out on a tight hammock of cords. The cotton used for fine spinning was also carefully washed; and even when it was not washed it was soaked with water and partially dried so that the fibres might be made to cling together.22 Most of the weaving was done by men, and until the invention of the fly-shuttle they cast the shuttle from hand to hand in the manner of their remotest ancestors. For the making of the broader fabrics two weavers were required when the width was greater than the easy stretch of a man’s arms. Sometimes cloths were woven wide and then split into two or more: hence the term “splits.” This became a common practice when the hand-loom workers were groaning under the pressure of competition from the power-loom.

We now reach the era of the great inventions. In order to ensure clearness it will be desirable to consider separately the branches of spinning and weaving: to pass from the one to the other, and follow the chronological order, The invention of machinery. might cause confusion. First emphasis must be laid upon the point that it was not mechanical change alone which constituted the industrial revolution. No doubt small hand-looms factories would have become the rule, and more and more control over production would have devolved upon the factory master, and the work to be done would have been increasingly assigned by merchants, had the steam-engine remained but the dream of Watt, and semi-automatic machinery not been invented. The spirit of the times was centralizing management before any mechanical changes of a revolutionizing character had been devised. Loom-shops, in which several journeymen were employed, were not uncommon: thus “in the latter part of the last (18th) and the beginning of the present (19th) century,” says Butterworth, describing the state of affairs in Oldham and the neighbourhood, “a large number of weavers ... possessed spacious loom-shops, where they not only employed many journeymen weavers, but a considerable proportion of apprentice children.” It is true that both the fly-shuttle and drop-box had been invented by that time, but the loom was still worked by human power. Specialism, however, was on the increase, the capitalist was assuming more control, and the operative was being transformed more and more into the mere executive agent. Further, as creative of enterprise, an atmosphere of freedom and a general economic restlessness, consequent upon the reaction against mercantilism, were noticeable. Great changes, no doubt, would soon have swept over Lancashire had a new source of power and big factories not been rendered essential by inventions in spinning.

The chief inventors were Lewis Paul and John Wyatt, James Hargreaves and Samuel Crompton. The two first originated the principle of spinning by rollers. Their patent was taken out in 1738, but no good came of it immediately, though Spinning and preparatory machinery. many trials were made and moderately large sums of money were lost. Ultimately Richard Arkwright brought forward the same plan improved:23 his first patent was dated 1769. Over the real authorship of the fundamental idea there has been much controversy, and it has not been absolutely proved that the second inventor, whether Thomas Highs, Arkwright or John Kay (a clockmaker of Warrington who assisted Arkwright to construct his machine and is said by some to have told him of an invention by Highs), did not hit upon the device afresh in ignorance of the work already done. Even as between Paul and Wyatt it is not easy to award due measure of praise. Probably the invention, as a working machine, resulted from real collaboration, each having an appreciable share in it. Robert Cole, in his paper to the British Association in 1858 (reprinted as an appendix to the 1st ed. of French’s Life of Crompton), championed the claims of Paul, but Mantoux, in his La Révolution industrielle au XVIIIe siècle, after studying the Wyatt MSS., inclines to attribute to Wyatt a far more important position, though he dissents from the view of Baines, who ascribes little or nothing to Paul.

Arkwright’s prospects of financial success were much greater than those of his predecessors, because, first, there was more need in his time of mechanical aids, and secondly, he was highly talented as a business man. In 1775 he followed up his patent of 1769 with another relating to machinery for carding, drawing and roving. The latter patent was widely infringed, and Arkwright was compelled to institute nine actions in 1781 to defend his rights. An association of Lancashire spinners was formed to defend them, and by the one that came to trial the patent was set aside on the ground of obscurity in the specifications. Arkwright again attempted to recover his patent rights in 1785, after the first patent had been in abeyance for two years. Before making this further trial of the courts he had thought of proceeding by petition to parliament, and had actually drawn up his “case,” which he was ultimately dissuaded from presenting. In it he prayed not only that the decision of 1781 should be set aside, but that both patents should be continued to him for the unexpired period of the second patent, i.e. until 1789. In his “case” (i.e. the petition mentioned above) Arkwright stated that he had sold to numbers of adventurers residing in the different counties of Derby, Leicester, Nottingham, Worcester, Stafford, York, Hertford and Lancaster, many of his patent machines, and continued: “Upon a moderate computation, the money expended in consequence of such grants (before 1782) amounted to at least £60,000. Mr Arkwright and his partners also expended in large buildings in Derbyshire and elsewhere upwards of £30,000, and Mr Arkwright also erected a very large and extensive building in Manchester at the expense of upwards of £4000. Thus a business had been formed which already (he calculated) employed upwards of five thousand persons, and a capital on the whole of not less than £200,000.”24 It is impossible to discover exactly the rights of the matter. Certainly Arkwright had been intentionally obscure in his specifications, as he admitted, and for his defence, namely that it was to preserve the secret for his countrymen, there was only his word. He may have hoped to keep the secret for himself; and as to the originality of both inventions there were grave doubts. But Arkwright has received little sympathy, because his claims were regarded as grasping in view of the large fortune which he had already won. He began work with his first partners at Nottingham (when power was derived from horses) and started at Cromford in 1771 (where the force of water was used). Soon he was involved in numerous undertakings, and he remained active till his death in 1792. He had met throughout with a good deal of opposition, which possibly to a man of his temperament was stimulating. Even in the matter of getting protective legislation reframed to give scope to the application of the water-frame, a powerful section of Lancashire employers worked against him. This protective legislation must here be shortly reviewed.

In 1700 an act had been passed (11 & 12 William III. c. 10) prohibiting the importation of the printed calicoes of India, Persia and China. In 1721 the act 7 George I. c. 7 prohibited the use of any “printed, painted, stained or dyed calico,” excepting only calicoes dyed all blue and muslins, neckcloths and fustians. This act was modified by the act 9 George II. c. 4 (allowing British calicoes with linen warps). Thus the matter stood as regards prints when Arkwright had demonstrated that stout cotton warps could be spun in England, and at the same time the officers of excise insisted upon exacting a tax of 6d. from the plain all-cottons instead of the 3d. paid by the cotton-linens, on the ground that the former were calicoes. Arkwright’s plea, however, was admitted, and by the act 14 George II. c. 72 the still operative part of the act of 1721 was set aside, and the manufacture, use, and wear of cottons printed and stained, &c., was permitted subject to the payment of a duty of 3d. per sq. yd. (the same as the excise on cotton-linens) provided they were stamped “British manufactory.” The duty was varied from time to time until its repeal in 1832.

Some more powerful force than that of man or horse was soon needed to work the heavy water-frames. Hence Arkwright placed his second mill on a water-course, fitting it with a water-wheel, and until the steam-engine became economical most of the new twist mills were built on water-courses. On rare occasions the old fire-engines seem to have been tried.

The following passage quoted from a note in Barnes’s History illustrates the pressing need of the early mills: “On the river Irwell, from the first mill near Bacup, to Prestolee, near Bolton, there is about 900 ft. of fall available from mills, 800 of which is occupied. On this river and its branches it is computed that there are no less than three hundred mills. A project is in course of execution to increase the water-power of the district, already so great and so much concentrated, and to equalize the force of the stream by forming eighteen reservoirs on the hills, to be filled in times of flood, and to yield their supplies in the drought of summer. These reservoirs, according to the plan, would cover 270 acres of ground, and contain 241,300,000 cub. ft. of water, which would give a power equal to 6600 horses. The cost is estimated at £59,000. One reservoir has been completed, another is in course of formation, and it is probable that the whole design will be carried into effect.”25

As early as 1788 there were 143 water-mills in the cotton industry of the United Kingdom, which were distributed as follows among the counties which had more than one.26

Lancashire 41 Flintshire 3

Derbyshire 22 Berkshire 2

Nottinghamshire 17 Lanarkshire 4

Yorkshire 11 Renfrewshire 4

Cheshire 8 Perthshire 3

Staffordshire 7 Midlothian 2

Westmorland 5 Isle of Man 1

The need of water to drive Arkwright’s machinery, and its value for working other machinery, caused a strong decentralizing tendency to show itself in the cotton industry at this time, but more particularly in the twist-spinning branch. Ultimately the steam-engine (first used in the cotton industry in 1785) drew all branches of the industry into the towns, where the advantages of their juxtaposition—i.e. the external economies of centralization—could be enjoyed. Out of the crowding of the mills in one locality sprang the business specialism which has continued up to the present day. Here it will not be out of place to notice the appearance of the new power, electricity, in the cotton industry, the extension of which may involve striking economic changes. The first electric-driven spinning-mill in Lancashire, that of the “Acme” Spinning Company at Pendlebury, the work of which is confined to the ring-frame, was opened in 1905. Power is obtained from the stations of the Lancashire Power Company at Outwood near Radcliffe, some 5 m. distant.

The chief principle of the water-frame was the drawing out of the yarn to the required degree of tenuity by sets of gripping rollers revolving at different speeds. This principle is still applied universally. Twist was given by a “flyer” revolving round the bobbin upon which the yarn was being wound; the spinning so effected was known as throstle-spinning. The plan is still common in the subsidiary processes of the cotton industry, but for spinning itself the ring-frame, which appears to have been invented simultaneously in England and the United States (the first American patent is dated 1828), is rapidly supplanting the throstle-frame,27 though the “ooziness” of mule yarn has not yet been successfully imitated by ring-frame yarn. The great invention relating to weft-spinning was the jenny, introduced by James Hargreaves probably about 1764, and first tried in a factory four years later.28 Hargreaves unfortunately was unable to maintain his patent, because he had sold jennies before applying for protection. Crompton’s mule, which combined the principles of the rollers and the jenny, was perfected about 1779. Both jennies and mules were known as “wheels,” because they were worked in part by the turning of a wheel. As they could be set in motion without using much power, being light when of moderate size, for a long time they were worked entirely by hand or partially with the aid of horses or water. The first jenny- and mule-factories were small for this reason, and also because skill in the operative was a matter of fundamental importance,29 as it was not in twist-spinning on the water-frame. The size of the typical weft-spinning mill suddenly increased after the scope for the application of power was enlarged by the use of the self-actor mule, invented in 1825 by Richard Roberts, of the firm of Sharp, Roberts & Co., machinists, of Manchester. In 1830 Roberts improved his invention and brought out the complete self-actor. Self-actors had been put forward by others besides Roberts—for instance by William Strutt, F.R.S. (son of Arkwright’s partner), before 1790; William Kelly, formerly of Lanark mills, in 1792; William Eaton of Wiln in Derbyshire; Peter Ewart of Manchester; de Jongh of Warrington; Buchanan, of Catrine works, Scotland; Knowles of Manchester; and Dr Brewster of America30—but none had succeeded. And Roberts’s machines did not immediately win popularity. For a long time the winding done by them was defective, and they suffered from other imperfections. Broadly speaking, until the American Civil War the number of hand-mules in use remained high. It was for the fine “counts” in particular that many employers preferred them.31 About the end of the ’sixties, however, and in the early ’seventies, great improvements were effected in machinery, partly under the stimulus of a desire to elevate its fitness for dealing with short-staple cotton, and it became evident that hand-mules were doomed. Here we may suitably refer to the scutching machine for opening and cleaning cotton, invented by Mr Snodgrass of Glasgow in 1797, and introduced by Kennedy32 to Manchester in 1808 or 1809; the cylinder carder invented by Lewis Paul and improved by Arkwright; and the lap-machine first constructed by Arkwright’s son.

We now transfer our attention to that accumulation of improvements in manufacturing (as weaving is technically termed) which, taken in conjunction with the inventions already described, presaged the large factory system which Weaving machinery. covers Lancashire to-day. Gradually, for many years, the loom had been gathering complexities, though no fundamental alteration was introduced into its structure until 1738, when John Kay of Bury excited the wrath of his fellow-weavers by designing and employing the device of the fly-shuttle. For some unfathomable reason—for the opposition of the weavers hardly explains it, though they expressed their views forcibly and acted upon them violently—this invention was not much applied in the cotton industry until about a quarter of a century after its appearance. The plan was merely to substitute for human hands hammers at the ends of a lengthened lathe along which the shuttle ran, the hammers being set in motion by the jerking of a stick (the picking peg) to which they were attached by strings. The output of a weaver was enormously increased in consequence. In 1760 John Kay’s son Robert added the drop-box, by the use of which many different kinds of weft could be worked into the same fabric without difficulty. It was in fact a partitioned lift, any partition of which could be brought to a level with the lathe and made for the time continuous with it. The drop-box usefully supplemented the “draw-boy,” or “draught-boy,” which provided for the raising of warps in groups, and thereby enabled figured goods to be produced. The “draw-boy” had been well known in the industry for a long time; in 1687 a Joseph Mason patented an invention for avoiding the expense of an assistant to work it,33 but there is no evidence to show that his invention was of practical value. Looms with “draw-boys” affixed, which could sometimes be worked by the weavers themselves, later became common under the name of harness-looms, which have since been supplanted by Jacquard looms, wherein the pattern is picked out mechanically.

The principle of the fly-shuttle was a first step towards the complete mechanizing of the action required for working a loom. The second step was the power-loom, the initial effort to design which was created by the tardiness of weaving as contrasted with the rapidity of spinning by power. After the general adoption of the jenny, supplies of yarn outran the productive powers of the agencies that existed for converting them into fabrics, and as a consequence, it would seem, some yarn was directed into exports which might have been utilized for the manufacture of cloth for export had the loom been more productive. The agitation for the export tax on yarn at the end of the 18th, and in the first years of the 19th century, is therefore comprehensible, but there was no foundation for some of the allegations by which it was supported. For a large proportion of the exported yarn, fabrics could not have been substituted, since the former was required to feed the hand-looms in continental homes and domestic workshops, against much of the product of which there was no chance of competing. The hand-loom was securely linked to the home of the peasant, and though he would buy yarn to feed his loom he would not buy cloth and break it up.34

Cartwright’s loom was not the first design adapted for weaving by power. A highly rudimentary and perfectly futile self-actor weaving machine, which would have been adapted for power-working had it been capable of working at all, had been invented by a M. de Gennes: a description of it, extracted from the Journal de sçavans, appeared in the Philosophical Transactions for July and August 1678, and again in the Gentleman’s Magazine in 1751 (vol. xxi. pp. 391-392). It consisted of mechanical hands, as it were, that shot in and out of the warp and exchanged the shuttle.35 Another idea, which however proved fruitful, was that of grinding the shuttle through the warps by the agency of cog-wheels working at each end upon teeth affixed to the upper side of the shuttle. Though shuttles could not in this fashion be set in rapid movement, the machine turned out to be economical for the production of ribbons and tapes, because many pieces could be woven by it at once. These contrivances were known as swivel-looms, and in 1724 Stukeley in his Itinerarium curiosum wrote that the people of Manchester have “looms that work twenty-four laces at a time, which was stolen from the Dutch.” Ogden says also that they were set up in imitation of Dutch machines by Dutch mechanics invited over for the purpose. Another interesting passage relating to the swivel-looms will be found in the rules of the Manchester small-ware weavers dated 1756, where the complaint is made that the masters have acquired by the employment of “engine or Dutch looms such large and opulent fortunes as hath enabled them to vie with some of the best gentlemen of the country,” and it is alleged that these machines, which wove twelve or fourteen pieces at once, “were in use in Manchester thirty years ago.”36 One power-factory at least was devoted to them as early as 1760, namely that of a Mr Gartside at Manchester, where water-power was applied, but the enterprise failed.37 Cartwright’s invention was probably perfected in its first form about 1787, but many corrections, improvements and additions had to be effected before it became an unqualified success. Cartwright’s original idea was elaborated by numerous followers, and supplementary ideas were needed to make the system complete. Of the latter the most important were those due to William Radcliffe, and an ingenious mechanic who worked with him, Thomas Johnson, which were patented in 1803 and 1804. They related to the dressing of the warp before it was placed in the loom, and for the mechanical taking up of the cloth and drawing forward of the warp, so that the loom had not to be stopped for the cloth to be moved on and the warp brought within play of the shuttle to be sized. Looms fitted with the latter of these devices were known as “dandy” looms. The looms that followed need not be described here, nor need we concern ourselves with the degree in which some were imitations of others. It is of interest to note, however, in view of recent developments, that one of Cartwright’s patents included a warp-stop motion, though it was never tried practically so far as the writer is aware. Looms with warp-stop motions are now common in the United States, as are also automatic looms, but both are still the exception in Lancashire for reasons that will be sketched later.

Power-looms won their way only very gradually. Cartwright and others lost fortunes in trying to make them pay, but the former was compensated by a grant of £10,000 from government. In 1813 there were 2400 only in the whole of the United Kingdom; in 1820 there were 14,000, beside some 240,000 hand-looms; in 1829, 55,500; in 1833, 100,000; and in 1870, 440,700.38 To-day there are about 700,000 in the cotton industry. The beginning, and the final consequences, of the competitive pressure of the power-looms may be read in the reports of official inquiries and in Rowbotham’s diary.39 It was upon the fine work that the hand-loom weavers retained their last hold. In 1829 John Kennedy wrote in his paper to the Manchester Literary and Philosophical Society on “The Rise and Progress of the Cotton Trade,” “It is found ... that one person cannot attend upon more than two power-looms, and it is still problematical [even in 1829, observe] whether the saving of labour counterbalances the expense of power and machinery and the disadvantage of being obliged to keep an establishment of power-looms constantly at work.” It was not easy to obtain a sufficiency of good hands for the power-looms, because the operatives, who had acquired their habits under the domestic system, hated factory life. This, in conjunction with the ease with which the art of coarse weaving could be acquired and the cheapness of rough looms, helps to explain the wretched straits into which the hand-loom weavers were driven.

Improvements in machinery, which ultimately affected every process from cleaning the cotton to finishing the fabric, and the application of water and steam-power, so lowered the cost of production as to render Lancashire the cotton Growth. factory of the world. Figures are quoted in the table to show the rate of growth in different periods of England’s imports and exports as regards the raw material and products of this industry. It is important to remember when reading the last 6 columns that the value of money was the same in 1831-1835, 1851-1855 and 1876-1880: the sums of Sauerbeck’s index numbers for these periods were 454, 451 and 444 respectively. In the last two periods there were considerable depressions in prices. If prices had remained constant, in the periods 1891-1895 and 1896-1900 the figures of exports would have been £90 millions and £91 millions respectively. The growth in trade has been partly occasioned by the enormous increase in the volume of cotton goods consumed all over the world, which in turn has been due to (1) the growth of population, (2) the increase in productive efficiency and well-being, and (3) the substitution of cotton fabrics for woollen and linen fabrics. The rate of growth between the periods 1771-1781 and 1781-1791 (which is not shown in the above table) was particularly remarkable, and reached as high a figure (when measured by importations of weight of cotton) as 320%.

Year. Imports of Raw Cotton, Million ℔. Raw Cotton re-exported, Million ℔. Exports of Cotton Yarns and Manufactures, Million £. Imports of Cotton Yarns and Manufactures, Million £

Yarns. Manu- factures. Total. Yarns. Manu- factures (excluding Lace). Total.

1700-1705 1.17 · · · · · · · · · · · · · ·

1771-1775 4.76 · · · · · · · · · · · · · ·

1785-1789 · · · · · · · · 1.07* · · · · · ·

1791-1795 26.00 · · · · · · 2.09* · · · · · ·

1816-1820 139.00 10.00 2.5 13.8 16.30 · · · · · ·

1831-1835 313.00 23.00 4.8 14.2 19.00 · · · · · ·

1851-1855 872.00 124.00 6.8 24.9 31.70 · · · · · ·

1876-1880 1456.00 180.00 12.4 56.1 68.30 · · 2.29 2.29

1891-1895 1746.00 217.00 9.7 56.6 66.30 .42 2.78 3.20

1896-1900 1798.00 223.00 8.9 58.2 67.10 .26 4.27 4.53

1901-1905 1920.00 265.00 8.4 70.7 79.10 .22 5.10 5.32

* Official values.

Nothing is more interesting in the cotton industry than the processes of differentiation and integration that have taken place from time to time. Weaving and spinning had been to a large extent united in the industry in its earliest form, Differentiation and Integration. in that both were frequently conducted beneath the same roof. With mechanical improvements in spinning, that branch of the industry became a separate business, and a substantial section of it was brought under the factory régime. Weaving continued to be performed in cottages or in hand-loom sheds where no spinning at all was attempted. Cartwright’s invention carried weaving back to spinning, because both operations then needed power, and the trouble of marketing yarn was largely spared by the reunion. Mr W. R. Grey stated in 1833 to the committee of the House of Commons on manufactures, commerce and shipping, that he knew of no single person then building a spinning mill who was not attaching to it a power-loom factory. Some years later the weaving-shed split away from spinning, partly no doubt because of the economies of industrial specialism, partly because of commercial developments, to be described later, which rendered dissociation less hazardous than it had been, and partly because, in consequence of these developments, much manufacturing (as weaving is termed) was constituted a business strikingly dissimilar from spinning. The manufacturer runs more risks in laying by stocks than the spinner, because of the greater variety of his product and the more frequent changes that it undergoes. The former, therefore, must devote more time than the latter to keeping his order book and the productive power of his shed in close correspondence. The minute care of this kind that must be exercised in some classes of businesses explains why the small manufacturer still holds his own while the small spinner has been crushed out. It also explains to some extent the prevalence of joint-stock companies in spinning, and their comparative rarity in manufacturing. Here we should notice, perhaps, that the only combination of importance in the cotton industry proper (apart from calico-printing, bleaching, &c., and the manufacture of sewing-cotton) is the Fine Cotton Spinners and Doublers Association, founded in 1898, which is practically coextensive with fine spinning and doubling.

The specialism of the two main branches of the industry has been followed by the specialism of sub-branches and by the localization of specialized parts. Of the localization of certain sections of the cotton industry Localization of branches of the industry. the late Mr Elijah Helm, who spoke with the authority of great local knowledge, has written as follows:—

“Spinning is largely concentrated in south Lancashire and in the adjoining borderland of north Cheshire. But even within this area there is further allocation. The finer and the very finest yarns are spun in the neighbourhood of Bolton, and in or near Manchester, much of this being used for the manufacture of sewing-thread; whilst other descriptions, employed almost entirely for weaving, are produced in Oldham and other towns. The weaving branches of the industry are chiefly conducted in the northern half of Lancashire—most of it in very large boroughs, as Blackburn, Burnley and Preston. Here, again, there is a differentiation. Preston and Chorley produce the finer and lighter fabrics; Blackburn, Darwen and Accrington, shirtings, dhooties and other goods extensively shipped to India; whilst Nelson and Colne make cloths woven from dyed yarn, and Bolton is distinguished for fine quiltings and fancy cotton dress goods. These demarcations are not absolutely observed, but they are sufficiently clear to give to each town in the area covered by the cotton industry a distinctive place in its general organization.”40

The present local distribution of the cotton industry, as far as it is displayed statistically, is revealed in the table beneath, based upon the figures of spindles and looms given by Worrall and those of operatives in the census returns of 1901.

Distribution of Cotton Operatives in Lancashire and the Vicinity according to the Census Returns of 1901, together with the Number of Spindles and Looms according to Worrall.

No. of Operatives. No. of Spindles (in Thousands). No. of Looms.

Blackburn 41,400 1,325 75,300

Bolton 29,800 5,035 20,100

Oldham 29,500 11,603 18,500

Burnley 27,900 687 79,300

Manchester and Salford 27,200 2,666 24,200*

Preston 25,000 2,036 57,900

Rochdale 14,800 2,168 25,100

Darwen 12,500 336 28,700

Nelson 12,400 23 39,000

Glossop** · · 968 15,400

Bury 10,700 818 22,200

Stockport 9,700 1,803 8,700

Ashton-under-Lyne 8,600 1,839 11,500

Accrington 8,300 417 36,400

Colne 7,300 140*** 20,500

Heywood 7,300 869 6,400

Stalybridge 7,100 1,106 7,100

Todmorden 6,900 261 15,800

Rawtenstall 6,600 356 8,800

Hyde 6,500 553 7,900

Chadderton 6,400 · · · ·

Haslingden 6,100 148 12,000

Bacup 5,900 315 9,300

Chorley 5,900 547 17,900

Farnworth, near Bolton 5,700 738 10,600

Leigh 5,000 1,667 5,900

Great Harwood 4,900 72 12,400

Middleton 4,900 511 2,500

Radcliffe 4,800 157 8,900

* Manchester only.

** The number of operatives in places in Derbyshire is not separately specified.

*** Includes Foulridge with Colne.

Local markets have steadily lost in importance, partly owing to railway development, and it is now almost entirely in Manchester, on the Exchange, that dealing in yarns and fabrics takes place, and arrangements are made for export. The old Manchester Exchange, built in 1729, was taken down in 1792. A new Exchange, reared on a contiguous site, was opened in 1809, the first stone having been laid in 1806. The present building was erected in 1869. The great bulk of the exports of cotton goods proceeds from Liverpool, though London used to be the leading port, and Liverpool is still the chief English market for raw cotton, though now from one-sixth to one-eighth of English cotton supplies come up the Manchester Ship Canal.

To understand the present organization of the cotton industry the reader must begin by mentally separating the commercial from the industrial functions. By the industrial functions are meant the arrangements of factors in Modern organization. production—choosing the most suitable machinery and hands, combining them in the most economical system, adapting the material used to this system, and keeping its working at the highest attainable level. The commercial functions consist in business which is not industrial. Analysis will show that there are, broadly speaking, two classes of commercial functions, namely (1) arranging for purchases and sales, and (2) the bearing of risks. The character of the former is apparent; it consists, as regards yarn, in discovering for each manufacturer which spinner makes the yarn which is best adapted to his requirements at the lowest cost, and in finding the most suitable customers for spinners. Risk-bearing is a commercial function of another kind. Every business that involves anticipation involves commercial risks. Thus the spinner who sells “forward” yarn, trusting that the price of cotton will not rise, is taking commercial risks, and so is the spinner who produces for stock, trusting that the class of yarn that he is making will continue in demand. These two instances will suffice to indicate what is meant by the carrying of commercial risks. To make the rest of our argument clear it will be well to write down formulae. Let A and B represent respectively the industrial operations of spinning and manufacturing. Let a and α represent respectively the commercial operations implied by the separate existence of A, that is, the buying of cotton and the selling of yarn; and let b and β stand for the commercial operations associated with manufacturing, that is, the buying of yarn on the one hand, and the finding of customers and arranging for their purchases on the other hand. Then, A and B being distinct businesses, it is obvious that a range of schemes is possible of which the extremes may be roughly represented as follows:—

1. (aAα), (bBβ) 2. (a), (A), (αb), (B), (β),

where the brackets signify independent businesses. In case 1 each spinning business would be engaged with three problems, namely, (i.) buying material at the most favourable time, (ii.) producing at the lowest cost, and (iii.) finding buyers and selling at the highest price, including the arranging for the performance of the most remunerative work. But in case 2 the spinner would confine his attention to purely industrial matters, while the problem of finding cotton and arranging for the bearing of the risks as to future prices would rest with other persons, and the business of bringing spinner and manufacturer together and taking such risks as may be involved in ordering or disposing of yarn would be the function of yet others. In case 2 the commercial functions may be said to have differentiated completely from the main body of the industry. We need hardly give illustrations of the intermediate arrangements that formally lie between cases 1 and 2. A may retain commercial risks but find customers through intermediaries; in such an event there would be only partial differentiation of the commercial functions. The reader must be reminded also that for the sake of simplicity in the formulae we have overlooked different classes of A and of B, omitted bleaching, dyeing, printing and finishing, and drawn no distinction between the various classes of commercial work covered by one letter, for instance, selling in the home market and selling abroad.

It may help the reader to appreciate the organic growth of the cotton industry if we now run over the main lines of its evolution. Originally the industrial units were held together in one homogeneous commercial setting. The Manchester merchants bought cotton and warps, put them out to the weavers, and arranged for the finishing of the cloth and then for its sale, so far as they had not been acting on orders already received. There were variations of this system—for instance, in early years weavers sometimes bought their own yarns and cotton and sold their cloth—but just before the industrial revolution the arrangement sketched above was the most usual. Adverting to our formula, the Manchester merchants, we observe, performed functions a (in conjunction with importers), b (as regarded warps), and β. Weft the weaver had to get spun by his family or outsiders. So, broadly speaking, there was one single commercial setting. After the appearance of the factory, the commercial work as between the water-twist mills, the mule-spinning businesses and the manufacturers, so far as the businesses were distinct, appears to have been done by the several producing firms concerned. It was not at once that (αb) began to differentiate, β was already a separate business in the hands of Manchester merchants and the foreign houses who had established themselves in Manchester to direct the export trade. At the present time an advanced stage of commercial specialism has been reached. From the risks connected with the buying of cotton the spinner may if he please escape entirely.41 Selling work is now done usually through intermediaries, but there is no one uniform rule as to the carrying of the commercial risks involved. This appears to be now to some extent a matter of arrangement between the persons concerned, but ultimately no doubt the risks will have to be borne by those most qualified by experience to bear them, namely, the commercial specialists. In no other trade in England, and in no other cotton industry abroad, has commercial specialism been carried so far as in the cotton trade of Lancashire. It is partly in consequence of the difference in this respect between the cotton industry in Lancashire and abroad that the separation of spinning from weaving is far more common in England than elsewhere. Elsewhere producers are deterred from specializing processes further in distinct businesses by the fear of the worries of buying and selling as between them.

The explanation of differences in respect of the degree of commercial specialism in different places and industries can be formulated only very generally. Time is required for the differentiation and localization to take place. The English cotton trade had not advanced very far in the “’thirties,” if we are to judge from the evidence given to commissions and parliamentary committees. The general conditions under which commercial specialism evolves may be taken to be a moderately limited range of products which do not present many varieties, and the qualities of which can be judged generally on inspection. In such circumstances private markets need not be built up, as they must be, for instance, for a new brand of soap which claims some subtle superiority to all others. Soaps under present conditions must be marketed by their producers. Broadly stated, if there be little competition as to substitutes, though there may be much as to price in relation to quality, commercial functions may specialize. On the whole this is the case in the cotton industry; in so far as it is not and firms produce specialities, they undertake much of the marketing work themselves.

The advantages of commercial specialism are numerous. Firstly it allows of differentiation of industrial processes, and this, of necessity, is accompanied by increasing returns. When weaving dissociates from spinning, both the number of looms in each business and the number of spindles in each business tend to increase; more division of labour is therefore secured, and lower costs of production are reached, and there is a further gain because producers concentrate their attention upon a smaller range of work. Again when producers are freed entirely, or to some extent, from commercial worries, they can attain a higher level of efficiency at the industrial task of mill organization, and a more perfect accommodation of capacity to function will be brought about. If the business unit is (aAα), a particular person may retain his place in the market by reason of his excellence at the work a or α, though as works organizer (i.e. at the performances of function A) he may be incompetent. The heads of businesses will succeed according to their average capacities at the three tasks a, A and α, and there is no guarantee, therefore, that any one of these tasks will be performed with the highest attainable efficiency in our present somewhat immobile economic system. But if the three functions are separated there is more certainty of a person’s success in the performance of each determining his continued discharge of it. The problems that arise when specialized markets become very highly developed are dealt with in the article Cotton: Marketing and Supply.

The distribution of cotton operatives among the chief centres has already been shown, but their distribution between processes has yet to be considered, and the proportions of different ages and sexes from time to time, together with the Operatives in various processes. total. With such statistical material as is available relating to supplies of labour we may set forth also the official returns made of the quantity of machinery at work from time to time. It hardly need be pointed out that the ratio of machinery to operatives roughly measures the efficiency of labour, other things being equal.

Machinery in the United Kingdom (in Thousands).

Years. Spinning Spindles. Doubling Spindles. Power- Looms.

1874 37,516 4366 463

1878 39,528 4679 515

1885 40,120 4228 561

1890 40,512 3993 616

1903 43,905 3952 684

Operatives employed in the Cotton Industry (in Thousands). (From the Census Returns.*) (The figures in italics relate to Married and Widowed Women.)

1901. 1891. 1881.

Lancashire. England and Wales. Lancashire. England and Wales. Lancashire. England and Wales.

M. F. M. F. M. F. M. F. M. F. M. F.

Cotton, card and blowing-room processes 11.4 28.7 13.8 34.0 · · · · · · · · · · · · · · · ·

10.1 12.2

Cotton spinning processes 49.5 19.6 64.1 28.6 · · · · · · · · · · · · · · · ·

4.3 6.0

Cotton weaving, warping, &c. 57.6 113.5 66.1 130.8 · · · · · · · · · · · · · · · ·

13.0 15.8

Cotton winding, warping, &c. 14.8 38.6 18.3 48.9 · · · · · · · · · · · · · · · ·

38.1 44.4

Total 133.3 265.9 162.3 320.7 178.2 281.8 213.2 332.8 150.7 249.8 185.4 302.4

Cotton workers in other processes or undefined 29.0 6.7 34.5 9.4 · · · · · · · · · · · · · · · ·

1.8 2.3

Tape, manufacturer dealer · · · · · · · · .47 .25 .9 1.5 .4 .24 .7 1.2

Thread, manufacturer dealer · · · · · · · · .2 .9 .6 2.1 .1 .9 .5 1.7

Fustian, manufacturer dealer .6 1.2 2.1 2.6 1.1 2.9 3.2 5.0 1.7 3.5 3.0 5.2

.55 1.0

Cotton, calico, warehouseman, dealer · · · · · · · · · · · · · · · · 2.5 .3 3.2 .38

* Census classifications have been altered twice in the period covered by this table.

In Scotland there are less than 15,000 cotton operatives distributed as follows:—

In Thousands.

Card and blowing-room processes .4

Spinning-room processes 2.1

Winding, warping, &c. 2.7

Weaving, warping, &c. 6.8

Workers in other processes or undefined 2.8

——

Total 14.8

Operatives employed in Cotton Factories in the United Kingdom and Percentages of each Class. (From Returns of Factory Inspectors.)

1835. 1838. 1847. 1850. 1856. 1862. 1867.

Male and Female under 13, or half-timers. 13.2 45.7 5.8 4.6 6.5 8.8 10.4

Male, 13 to 18 12.5 16.6 11.8 11.2 10.3 9.1 8.6

Male, over 18 26.4 24.9 27.1 28.7 27.4 26.4 26.0

Female, over 13 47.9 53.8 55.3 55.5 55.8 55.7 55.0

Total number of Cotton Operatives 218,000 259,500 316,400 331,000 379,300 451,600 401,100

1870. 1874. 1878. 1885. 1890. 1895. 1901.

Male and Female under 13, or half-timers. 9.6 14.0 12.8 9.9 9.1 5.8 4.1

Male, 13 to 18 8.5 8.0 7.2 7.9 8.2 7.9 7.0

Male, over 18 26.0 24.1 25.3 26.4 26.9 27.6 27.8

Female, over 13 55.9 53.9 54.7 55.8 55.8 58.7 61.1

Total number of Cotton Operatives 450,100 479,600 483,000 504,100 528,800 538,900 513,000

Number of Operatives (in Thousands) engaged in Spinning, Manufacturing and Subsidiary Processes (excluding Lace-making, but including the Fustian Manufacture). (From Census Returns.)

Males. Females. Males and Females.

Under 15. 15-20. Over 20. All Ages. Under 15. 15-20. Over 20. All Ages. Under 15. 15-20. Over 20. All Ages.

1881 29 39 121 189 40 81 189 310 69 120 310 500

1891 36 45 137 218 50 94 197 341 86 139 334 560

1901 24 36 139 199 36 92 207 335 60 128 346 535

The fact that the branches of work covered by the figures are not identical explains discrepancies between this and the previous table.

Number of Operatives engaged in the Cotton Industry (Processes being distinguished and Ages and Sex). (From Special Returns made by Factory Inspectors.)

Males in Thousands. Females in Thousands. Total in Thousands.

Half- timers. Under 18. 18 and over. Half- timers. Under 18. 18 and over.

Spinning and Preparatory Processes

1896 5.58 22.24 71.44 4.40 30.12 78.96 212

1898-1899* 5.42 21.57 71.37 3.86 30.44 77.64 210

1901 4.98 21.10 68.98 3.10 30.98 81.68 211

Weaving and Preparatory Processes

1896 7.54 18.79 75.81 11.87 49.19 151.34 315

1898-1899* 6.21 17.29 72.74 10.38 48.38 150.99 306

1901 4.72 14.86 73.81 8.0 45.66 155.03 302

* Average for 1898 and 1899.

The figures in this table are not quite complete except for 1901; the relations between the changes shown for each class should nevertheless be accurately represented.

Index Numbers of Money, Wages and Prices.

1840. 1855. 1860. 1866. 1870. 1874. 1877. 1880. 1883. 1886. 1891. 1902.

Cotton operatives. 50 54 64 74 74 90 90 85 90 93 100 105

Average wages for eight trades 61 61 73 81 83 97 94 89 92 90 100 108.7*

Sauerbeck’s index number 103 73 99 102 96 102 94 88 82 69 72 69

Average price of wheat per quarter 66/4 40/3 53/3 49/11 46/11 55/9 56/9 44/4 41/7 31/- 37/- 28/1

* Average for a slightly different group.

Weekly Wages in the Manchester and District Cotton Trade.

1834. 1836. 1839. 1841. 1849. 1850. 1859. 1860. 1870. 1877. 1882. 1883. 1886.

s. d. s. d. s. d. s. d. s. d. s. d. s. d. s. d. s. d. s. d. s. d. s. d. s. d.

Spinners’ average 23 4 23 11 22 1 22 0 21 7 20 5 24 1 23 2 27 8 34 4 31 6 32 4 35 7

Big piecers’ average 11 0 9 3 8 6 8 8 8 6 13 0 10 0 10 0 11 0 12 4 16 0 16 0 13 7

Weavers’ average 11 0 10 2 9 6 9 6 10 6 10 3 11 2 10 8 12 2 15 1 15 6 15 0 13 3

The most noticeable features of these tables are the decrease in the proportion of children employed and the steady increase in the number of operatives as a whole until recent years. The contraction of the body of operatives of late years seems to have occurred primarily among children and young persons (where the first check would naturally be looked for), and secondarily among adult males. If allowance be made for the smaller value of children as compared with adults, and the census results be taken, it is not evident that there has been any diminution in the amount of labour-power; and if the factory inspectors’ returns be accepted, the falling off in the number of operatives cannot be proved to have taken place in either of the chief branches of the industry at so rapid a rate as to have occasioned the enforced dismissal of any hands. An industry which was not recruited at all would have dwindled at a greater rate. At least it may be inferred from these figures, when taken in conjunction with the large increase in spindles and looms, that the output per head has considerably advanced in spite of the rise in the average quality of both yarns and fabrics produced. This rise in the value per unit of the output accounts to some extent for the fact that wages have not been adversely affected of late.

Mr A. L. Bowley has calculated index numbers of wages for the leading trades, including the manufacture of cotton. Those for the cotton industry are given below, together with averages for cotton and wool workers, the building trades, mining, workers in iron, sailors, compositors Wages and piece-rate lists. and agriculturists (England), the numbers in each class being allowed for in the average. Side by side with these figures, Sauerbeck’s index numbers of general wholesale prices are given, together with the average prices of wheat per quarter.

It must be remembered that the figures given above for cotton workers and average wages for eight trades do not measure the differences between each, but only the differences between the movements of each. Actual average money wages in the cotton industry have probably been approximately those stated in the second table beneath, but as these figures are culled from various sources they must not be taken to indicate fluctuations.42

The wage of fine spinners exceeds the average wage of spinners by percentages varying from about 25 to 35. In the above figures the earnings of three classes of spinners are averaged.

The highest wages are earned by mule-spinners (who are all males); their assistants, known as piecers, are badly paid. Persons can easily be found, however, to work as piecers, because they hope ultimately to become “minders,” i.e. mule-spinners in charge of mules. The division of the total wage paid on a pair of mules between the minder and the piecers is largely the result of the policy of the spinners’ trade union. Almost without exception in Lancashire one minder takes charge of a pair of mules with two or three assistants according to the amount of work to be done. Among the weavers there is no rule as to the number of assistants to full weavers (who are both male and female), or as to the number of looms managed by a weaver, but the proportion of assistants is much less than in the spinning branches, perhaps because of the inferior strength of the weavers’ unions. For the calculation of wages piece-rate lists are universally employed as regards the payment of full weavers and spinners; some piecers get a definite share of the total wage thus assigned to a pair of mules, while others are paid a fixed weekly amount. Many ring-spinners are now paid also by piece-rate lists, and all other operatives are almost universally so paid, except, as a rule, the hands in the blowing-room and on the carding-machines. Spinning and weaving lists are most complicated; allowances are made in them for most incidents beyond the operatives’ control, by which the amount of the wage might be affected. Still, however, they could not cover all circumstances, and much is left to the manner of their application and private arrangement. They should be regarded as giving the basis, rather than as actually settling, the wage in all cases. The history of lists stretches back to the first quarter of the 19th century as regards spinners, and to about the middle of the century generally as regards weavers, though a weaving list agreed to by eleven masters was drawn up as early as 1834. There are still many different district lists in use, but the favourite spinning lists are those of Oldham and Bolton, and the weaving list most generally employed is that known as the “Uniform List,” which is a compromise between the lists of Blackburn, Preston and Burnley. Under the “Particulars Clause,” first included in a Factory Act in 1891 and given extended application in 1895, the particulars required for the calculation of wages must be rendered by the employer. As in spinning there used to be doubts about the quantity of work done, the “indicator,” which measures the length of yarn spun, is coming into general use under pressure from the operatives. We ought to observe here that the Oldham Spinning list differs from all others in that its basis is an agreed normal time-wage for different kinds of work on which piece-rates are reckoned. But in effect understandings as to the level of normal time-wages are the real basis everywhere. If the average wages in a particular mill are lower than elsewhere for reasons not connected with the quality of labour (e.g. because of antiquated machinery or the low quality of the cotton used), the men demand “allowances” to raise their wages to the normal level. Advances and reductions are made on the lists, and under the Brooklands Agreement, entered into by masters and men in the cotton spinning industry in 1893, advances and reductions in future must not exceed 5% or succeed one another by a shorter period than twelve months. The changes as a rule now are 5% or 2½%. In all branches of the cotton industry it is usual for a conference to take place between the interested parties before a strike breaks out, on the demand of one or other for an advance or reduction.

Organization among the workers in the cotton industry is remarkably thorough. Almost all spinners are members of trade unions, and though the weavers are not so strongly united, the bulk of them are organized. The piecers are admitted Trade Unions. as members of piecers’ associations, connected with the spinners’ associations and controlled by them. Attempts to form independent piecers’ unions have failed. Weavers’ assistants are included in the weavers’ unions, which may be joined in different classes, the benefits connected with which vary with the amounts paid. One subscription only, however, is imposed by each branch spinners’ association, but in all branches it is not the same, though every branch pays the same per member to the amalgamation. All the trade unions of the chief workers in the cotton industry are federated in the four societies: (1) the Amalgamated Association of Operative Cotton Spinners (created in 1853 and reformed in 1870), (2) the Northern Counties Amalgamated Association of Weavers (founded 1884), (3) the Amalgamated Association of Card and Blowing-room Operatives (established 1886), and (4) the Amalgamated Association of Power-loom Overlookers (founded 1884). These were not, however, the first attempts at federation, and the term “federation” must not be taken in any strict sense. The distribution of power between the central authority and the local Societies varies, but in some cases, for instance among the spinners, the local societies approximate as closely to the status of mere branches, as to that of independent units federated for limited objects. We ought also to mention the societies of warp-dressers and warpers, tape-sizers and cloth-workers and warehousemen. There is no one federation of all cotton-workers, but the United Textile Factory Workers has been periodically called into being to press the matter of factory legislation, and international textile congresses are occasionally held by the operatives of different countries.

As to employers, four extensive associations include almost all the organization among them, two concerned chiefly with spinning and two with weaving. The former two are the Federation of Master Cotton Spinners’ Associations with local associations and including 21,000,000 spindles, and the Bolton Master Cotton Spinners’ Association with 7,000,000 spindles; the latter two are the North and North-East Lancashire Spinners’ and Manufacturers’ Association, covering about 3,000,000 spindles in addition to a large section of the looms of Lancashire, and the United Cotton Manufacturers’ Association.43

Factory legislation began in the cotton industry, and in no industry is it now more developed. The first acts were those of 1802 and 1819, both of which applied only to cotton-mills, and the former of which related only to parish apprentices. Factory Acts. The first really important measure was that of 1833, which curtailed the abuse of child-labour, enforced some education and provided for factory inspectors, of whom there were at first only four. The next act of importance, that of 1844, was chiefly remarkable for its inclusion of all women among young persons. The proportion of women, young persons and children engaged in the cotton industry is so high, that most regulations affecting them, e.g. those relating to the hours of labour, must practically be extended to all cotton operatives. This act killed night work for “young persons,” and children were not allowed to work at night. The year 1847 saw the introduction of what was known as the Ten Hours Act—after the 1st of May 1848 the hours of young persons (women included) and children were not to exceed ten a day and fifty-eight a week. A further limitation of hours to 56½ a week was secured in 1874, and this was cut down by another hour (the concession of the 12 o’clock Saturday) in 1901. “Young persons” now includes all who are not half-timers and have not attained the age of eighteen, and all women. The rules as regards the employment of children, which have steadily improved, are at present as follows. No child under twelve may be employed. On attaining the age of thirteen the child may become a full-timer if he has obtained the prescribed educational certificate (i.e. fifth standard attainment or three hundred attendances each year for five consecutive years). Failing this he must wait till he is fourteen before he can be employed full time. Half-timers may be employed either (a) on alternate days, which must not be the same days in two successive weeks, or (b) in morning and afternoon sets. In the case of arrangement (a), the child when at work may be employed during the same period as a young person or woman, which in Lancashire is almost universally from 6 to 6 with two hours for meals.44 In the case of arrangement (b), which is the system generally adopted in Lancashire, a half-timer in the morning set works from 6 to 12.30, with half an hour for breakfast, and in the afternoon from 1.30 to 6 except on Saturdays, when the hours are from 6 till 11.30 for a manufacturing operative, or till 12 for other work, for instance, cleaning. The child must not work two consecutive weeks in the same set (that is, in mornings or afternoons), nor on two successive Saturdays, nor on Saturday at all if during any other day of the same week the period of employment has exceeded 5½ hours (i.e. a child in the morning set does not work on the Saturday). Other important features of factory legislation relate to the fencing of dangerous machinery and its cleaning when in motion (the regulations being strictest in the case of children and most lax in the case of male adults), and conditions of health, including the amount of steaming allowed, which was first regulated by the Cotton Cloth Factories Act of 1889.

The Cotton Industry outside England.

A brief survey will now be made of the cotton industry in parts of the globe other than the British Isles, and as a prelude the following broad estimates of the numbers of spindles and looms in the chief national seats of the cotton industry may be put forward.45 The table is further supplemented by other figures46 for the number of spindles at different times in the United Kingdom, the United States and the continent; and finally we may add the figures of cotton consumed.

The different average fineness of counts spun in different places must be borne in mind when the consumption of each district at the same time is being considered, but the relations between the amounts consumed in the contrasted districts in the two periods would not be affected much by this difference.

Estimated Population in 1902. In Millions. Million Spinning Spindles in 1909. Thousand Power- Looms about 1906.

United Kingdom. 42 53.5 700

United States 79 27.8 550

Germany 58 9.8 215

France 39 6.8 110

Russia 139 7.8 150

India 294 (1901) 5.8 45

Austria 26.7 4.2 80

Spain 18.6 (1900) 1.9 69

Italy 33 4.0 100

Switzerland 3.4 1.5 30

Japan 46 1.7 · ·

Belgium · · 1.2 · ·

Cotton Spindles (including Doubling Spindles) in Millions.

United Kingdom. Europe. United States. Other Countries. Total.

1870 37.7 13 7.1 · · 57.8

1880 44.5 21 10.6 2 78.1

1890 44.5 26 14.2 4 88.7

1900 46.2 32 19 7 104.2

1903 47.9 33 22.2 7.5 110.6

Average Annual Consumption of Cotton in the Period 1831-1835.

Millions of ℔.

United Kingdom 295

Continent of Europe 143

United States 79

Average Annual Consumption of Cotton in the Period 1900-1905.

Millions of ℔.

United Kingdom 1634

Continent of Europe 2486

United States 1995

Roughly the consumption of cotton per spindle in the three areas to-day is, in ℔, 35 for the United Kingdom, 70 for the continent, and 95 for the United States.

Before the cotton industry in other countries is described it will be necessary to explain how it could have developed there on a large scale at all. Of course this growth is to be accounted for very largely by the natural protection of cost of transport aided by tariffs. But it would be a mistake for Englishmen to imagine that all foreign cotton mills are the product of a forcing culture, and that if the favourable conditions created by import duties were removed they would totally disappear. No doubt some of the growth is artificial, but much is natural and would have taken place under universal free trade conditions. Much of it, indeed, would have appeared in these circumstances even were cost of production a negligible quantity, difficult though it may be at first to reconcile this statement with certain ordinary conceptions of the operations of the law of increasing returns. Lancashire secured an immense lead at the beginning of the 19th century, and if the cost of production may be represented as varying inversely as the magnitude of the industry, every addition to her success increased her advantages. How could the small industry, with a high cost of production because it was small, compete with Lancashire? The answer is to be found in the peculiar conditions governing international trade and a closer analysis of “increasing returns.” “Increasing returns” in any place are a function of two variables, (1) the magnitude of the world market under conditions of world commerce, and (2) the magnitude of the industry in the spot in question. The economies connected with the first variable, which in such an industry as the cotton industry are enormous, and govern ultimately the limits of business specialism, are shared by every national section of the industry whether it be great or small. If Haiti started a cotton factory she might import all her specialized machinery—the specialism involved in producing which is dependent upon the exportation of some of it—and restrict narrowly the work undertaken by her one factory. The cotton goods outside this range she would still import, and if her specialized product were in excess of local demand she could export some of it, if she were favourably placed in respect of cost of carriage, for cost of production in Haiti would not be impossibly high, since machinery and the general system of production would be quite up to date though labour might be highly inefficient. Of course, the country with a large industry enjoys high local economies, and it might be thought that these alone would be a menace to the stability of the small industry, because if the industry in the favoured locality increased these would increase also and the small industry would be undersold. The answer to this difficulty is that foreign trade depends upon ratios between ratios, that is, upon the ratios between the costs of production of all the products of each country in relation to similar ratios for other countries. Relatively, therefore, diminishing returns operate in every country. In every country there must come a time, the utility of commodities being taken into account, when a unit of labour and capital provides less utility when applied to the creation of cotton goods, say, than when applied to producing something else for home consumption or for export in exchange for commodities wanted at home. It becomes apparent, therefore, that cotton industries of widely varying sizes dispersed throughout the world can settle into relations of perfectly stable equilibrium, as that term is understood by the economist. Slow changes, of course, in their relative volumes might be looked for with changes in a mutable world, but very sudden collapses would be impossible unless the general course of human affairs were revolutionized.

The United States.—The machine-cotton industry was carried to North America almost as soon as it evolved in England. Models of Arkwright’s machines were smuggled across the Atlantic in 1786—Arkwright’s first mill had not been started in England until 1769—and these with a jenny and stock-card were publicly exhibited. From these models a great mass of machinery was soon constructed. The first mill was erected in 1788 (that of the Beverly Association), the second appeared in 1790, the third five years later, and in 1798 Samuel Slater started with some of his wife’s relatives the first mill in which the principle of the water-frame was carried throughout. It is said that it was not until 1814 that power-loom manufacturing was commenced, but in England success with the power-loom was long delayed. As early as 1831, however, there were in the United States—mainly in the New England states—800 factories, a million and a quarter spindles, 33,500 looms and 62,200 operatives. At this time the annual consumption of cotton was about 77,000,000 ℔ as compared with some 300,000,000 ℔ in England at the same date, and 2,000,000,000 approximately in the United States at the present time.47 Writing in 1840, James Montgomery said that, in respect of cost of production, the American industry was 19% behind that of England apart from the cost of raw material, which was then a good deal less to the Americans. In 1878, when there was much interest in the question of British efficiency in the cotton industry because the passage of the Factory Act of 1874 had cut down the working hours, the Economist contrasted the result of twenty-five years’ growth in England and America:—

“In 1853 the average English production per weaver of 8¼ ℔ shirting was 825 yds. per week of sixty hours. In 1878 the working hours had fallen to fifty-seven, and the production had risen to 975 yds. An increased production of 23% is thus due to improvement in the processes of manufacture. In 1865 there were 24,151 persons employed in Massachusetts in the production of cotton goods, and they produced 175,000,000 yds. In 1875 the operatives numbered 60,176, and their product was 874,000,000 yds. The operatives had increased 150% and their products had increased 500%. The increase of production due to improved methods was thus in England 23%, and in Massachusetts 100%. I do not, of course, suppose that the American manufacturer is in advance of his English rival to the extent of this difference, for I presume that he started upon the career of improvement from a lower platform. But a progress so greatly more rapid than ours will be admitted to cast much light on the change which has occurred in our relative positions.”

The contrast no doubt was not perfect, as indeed it could not be in view of the varieties of product and their changes, but it proves at any rate that Americans were making vast strides in industrial efficiency even before the period when American methods and American enterprise were monopolizing in a wonderful degree the attention of the business world.48 About a dozen years later the low real cost of production of simple fabrics in the United States was universally admitted, and also that American manufacturers were making more use of machinery than their European rivals. In a typical weaving shed in Massachusetts, for instance, of which particulars were published, twenty women “tended” as many as eight looms apiece, forty-three managed seven, two hundred and thirty-two managed six, and only eleven had five only.49 Since then, moreover, advance has been rapid, and the sudden development of the South has astonished the business community of other centres of the cotton industry.

Before the lines of development in America are specifically dealt with, and particularly the industrial phenomena in the South, a few words must be said of the general extension of the industry. The consumption of cotton in the United States in million ℔ was about 75 in 1830, 390 in 1860, 1100 in 1890 and nearly 2000 on an average of the five crop years from 1900-1901 to 1904-1905: active spindles advanced from 1,250,000 in 1830 to 10,653,000 in 1880 and about 21,250,000 in 1905. Looms which numbered 33,500 in 1830 had reached 226,000 in 1880 and nearly 550,000 in 1905. At the same time population, it must be remembered, was growing at a phenomenal rate: from 31.4 millions in 1860 it had passed to 38.6, 50.2, 62.6 and 76.3 at the succeeding decennial censuses, the decennial rates of increase being in order 22.5, 30, 25 and 20.5 as compared with 8.5, 10.5, 8 and 9 as shown by the corresponding censuses in the United Kingdom. Protection was of course contributory to the growth of the American cotton industry. It may be remarked incidentally that the New World, including the West Indies and the Chinese empire, take the bulk of American exports, which for so large an industry are inconsiderable. The imports have always been well in excess of the exports. The encouragement of home industries by tariffs was definitely aimed at after the war with England during the Napoleonic struggles, and although a sensible reduction of duties was experienced after 1845 the reaction to protection that followed the Civil War was never significantly departed from except by the single act of 1883. In 1790 the duties on cotton goods were 7½% ad valorem, and they rose gradually until they reached 25% in 1816. Slight reductions some seventeen years later were followed in the early ’forties by a tariff of 30%. Diminutions were succeeded by oscillations, though at no point was a low level touched. Severe charges were imposed in 1890, and after some relaxation in 1894 the policy of restrictiveness was restored in 1897. According to the calculations made by the English Board of Trade in 190350 no fabrics were admitted at a charge equivalent to less than 68% ad valorem, and no yarns were admitted at a charge lower than 45% ad valorem. Cotton thread is subjected to a rate equivalent to 375%51

The character of the growth of the cotton industry in the United States, as revealed by recent census returns, is peculiarly interesting:—

Thousands Percentage Increase

1880. 1890. 1900. 1905. 1880-1890 1890-1900 1900-1905

Active Spindles 10,653 14,188 19,008 23,156 33.8 34 21.8

Looms 226 325 451 541 43.90 38.7 20

℔ cotton consumed 750,344 1,117,946 1,814,003 1,875,075 48.99 62.3 3.3

Wages $42,041 $66,025 $85,126 $94,378 57 28.9 10.9

Capital $208,280 $354,021 $460,843 $605,100 70 30.2 31.3

Employees not officers and clerks 174.7 218.9 297.9 310.5 25.3 36.1 4.2

Cotton small wares are included in the totals for 1880 and 1890, but excluded from those for 1900 and 1905. We must observe further that “capital” is a vague term. Recent events in the United States afford a valuable empirical indication of the effect that improved machinery actually has upon wages. The new automatic looms caused a saving of labour per unit of product which recalled the complete subversion at the industrial revolution of the proportions in which the several factors in production were organized. Displacement of labour and falling wages might not unreasonably have been looked for temporarily, but wages stuck at their old level or rose. The rise was caused by numerous converging forces which brought their united weight to bear. First, prices so fell as the result of the new machinery that the increased volume of commodities which the market could absorb more than counterbalanced, it would seem, the labour-saving of the new machinery, the cotton industry being taken as a whole. It must be remembered that to increase the output from the subsidiary processes where labour had not been saved more hands had to be drafted in. Thus, a contraction of the body of weavers was accompanied by an expansion of the body of cotton operatives. Again weavers’ wages were naturally raised in a special degree because automatic machinery called for quick, trustworthy and intelligent hands, endowed with versatility, especially in the days when the machinery was still in the semi-experimental stage. The American employer tries to save in labour but not to save in wages, if a generalization may be ventured. The good workman gets high pay, but he is kept at tasks requiring his powers and is not suffered to waste his time doing the work of unskilled and boy labour. There is, certainly, in the American labour problem no serious grievance on the question of wages. If there is any abuse it consists in excessively fierce work. Mr. T. M. Young, who visited the American cotton districts in 1904 with an informal commission of Lancashire spinners and manufacturers, did not think that the cause of the high wages—allowance being made for the purchasing power of money, they are above those of England, though cotton operatives in England are well paid relatively—was the superiority of the American cotton worker; neither did the representatives of the English cotton operatives who accompanied the Moseley Commission. As often as not “the cotton operative in the United States is a French Canadian, a German, an Italian, a Hungarian, an Albanian, a Portuguese, a Russian, a Greek, or an Armenian.” It is the extensive “exploitation” of machinery seemingly, together with the speed of work, which keep wages high, combined with the horizontal and vertical mobility of American labour, which prevents it from accumulating in pools, and causes streams of the best hands to be flowing continuously to other callings and places, and no insignificant proportion to climb the social ladder. The remainder naturally profit, for a local or trade congestion of labour is avoided, and the voluminous recruiting of enterprise by the intensified competition among employers keeps the demand for labour high.

One noticeable point in the table quoted above is that until recently cotton consumed increased much faster than the number of spindles. This might be explained in a variety of ways. Average counts remaining constant, the average speed of the spindle might have risen; or the latter remaining constant, counts might have been getting finer. Speeds have certainly gone up a good deal of late on some counts. And it is quite likely, too, that concentration on the manufacture of coarse goods for export, with stout warps to keep down the breakages and raise the output per loom, may be reckoned as one cause.

Despite the recent sensational growth in the South, the New England States still remain the most prominent seat of the American cotton industry. They contained in 1905 about 14 million spindles as compared with 7.7 millions in the South and West, and their relative possession of looms approaches, though it does not quite reach, the same proportion. The leading States in the South in order of importance are South Carolina, North Carolina, Georgia and Alabama, and in the North, first Massachusetts with an enormous lead, then, in order, Rhode Island, New Hampshire, Connecticut, Maine, New York, Pennsylvania, New Jersey. The bulk of the cotton industry in the North is contained within a small area. A circle around Providence, Rhode Island, of 30 m. radius includes, according to the twelfth census, nearly 7¼ million spindles,—there were only 58,500 spindles in this area in 1809. Of the chief towns Fall River stood first in 1900 in value output, and was followed in order by Philadelphia, New Bedford, Lowell, Manchester and Pawtucket. The climate of Fall River is very similar to that of English spinning districts. Its population in 1900 was 105,000, and of these only 14,600 were of American parentage. Of the remainder, 16,700 were English, 17,800 Irish, 29,600 French Canadians and about 5000 Portuguese. Among the rest of foreign parentage, Armenians, Russians and Italians are numerous. But Massachusetts is famous for the number of immigrants it attracts. It is almost incredible, but nevertheless a fact according to a recent statistical report, that in 1903 as many as 91% of the cotton operatives of the State were of foreign descent—chiefly French Canadian and Irish. In 1902 there were nearly 90 mills at Fall River with 3,000,000 spindles and 16,000 looms. The spindles amount to about one-third of all in Massachusetts, but Fall River’s share of the looms of the State is not large. The spindles exceed in number those possessed by any State except of course the one in which it is placed. In comparison with a great spinning town in England, nevertheless, Fall River does not appeal strongly to the English imagination. It has little over a quarter of the spindles of Oldham, or three-fifths of those of Bolton,—among English towns it would stand third, i.e. between Bolton and Manchester and Salford, which, in spite of the movement of spinning to the hills, still holds in England a leading place. The whole of Massachusetts, it is of interest to observe, has fewer spindles than Oldham, and only about half those of Oldham and Bolton together. Originally it was the river which attracted the mills to Fall River, and as the water-power available was almost inexhaustible, it was possible for the mills to congregate together and for a town to grow up. In England, when much of the industry was dependent for power upon water, decentralization was entailed, for the thin streams of Lancashire could not support more than two or three mills at most in proximity. Hence in England, after Watt’s steam-engine had succeeded, the economies of centralization led eventually to the desertion of the mills on the water-courses. But at Fall River the perfecting of the application of steam-power merely involved its use to supplement the water-power on the old site. The presence of water-power explains half the success of New England. In the six States 35% of all the power used is derived from water, and in the cotton-manufacturing of these States water provides 32.6% of the power. For industrial purposes generally the river most exploited is the Merrimac, upon which stand the leading cotton towns of Lowell, Lawrence and Manchester. Hitherto little has been done in the way of using water to generate electric power.52

The two most striking features of the American industry to-day are the introduction of the automatic looms, already briefly referred to, and the development of the South. The Northrop Loom Company has spent a fortune in pushing its loom on to the market. It has not hesitated to share risks, and it has run one “advertisement” mill at least, namely that at Burlington, Vermont, with 55,000 spindles and nearly 1300 looms. In this mill the labour-saving is shown by the following figures, the looms being of two sizes, 32 in. and 44 in. Of the former, 3 weavers run 18 each, 39 tend 16 each, only a few odd weavers tend less than 16, and learners even are at work on 8 to 11 each; on the latter, of 29 weavers 17 mind 16 looms each and 12 mind 12 (on stripped fabrics).53 Of course a high level of efficiency would be expected in this show mill. That American employers have readily been converted to a belief in the economy of the new machinery we are not astonished to learn in view of the American temperament, the intensity of competition among business leaders, and the prevailing spirit of adventure. Thousands of workable old looms have been scrapped, and probably at the present time there are 100,000 automatic looms running in the United States. No other country can point to a rate of substitution which approaches that in the United States. The causes, apart from the temperamental and social to which reference has already been made, are probably (1) that there is disagreement as to the present economy of automatic looms on many fabrics,54 (2) that Americans aim at frequency of renewal of plant, and avoid making their machinery so durable as to prove ultimately, perhaps, a handicapping inheritance, and (3) that a greater bulk of American work is appropriate for the new looms than of English or continental work. But automatic machinery is being used increasingly in Lancashire.55 And the operatives ultimately benefit. It is the half-developed machine, to which labour must actually be linked as an essential part, which is responsible for monotonous work and creates the dislike of mechanical aids.

Now we turn to the recent development of the Southern States. Never has an industry grown faster than that of the two Carolinas, Georgia and Alabama. Some of the earliest experiments with the machine industry were conducted in South Carolina, but from that time till the end of the 19th century nobody imagined the possibility of a great Southern expansion. In 1880 the South contained less than half a million spindles—i.e. about as many as Hyde, Middleton or Chorley, and one-twenty-third of the numbers in Oldham. Twenty years later they had increased twelvefold and the Southern States, in respect of the number of spindles, had taken precedence of Bolton. To-day probably about eight and a half millions might be counted. In addition there are some two hundred thousand looms, or nearly as many as in the three leading cotton-weaving towns of England—Burnley, Blackburn and Preston. The rapid oncoming of the South may also be traced by its consumption of cotton—which as an index, however, is not perfect. This on an annual average was, in thousand bales, 164, 269, 453, 717 and 1233 in each of the periods 1876-1880, 1881-1885, 1886-1889, 1891-1895 and 1895-1900 successively. The consumption since then, as compared with that of the Northern States, Great Britain and the European continent, has been as follows. It must be remembered that the consumption per spindle varies greatly from place to place.

Consumption of Cotton in Thousand Bales of about 500 ℔ each.

Southern States. Northern States. Total United States. Great Britain. Europe.

1900-1901 1583 1963 3546 3269 4576

1901-1902 2017 2066 4083 3253 4836

1902-1903 1958 1866 3824 3185 5148

1903-1904 1889 2046 3935 3017 5148

1904-1905 2270 2292 4562 3620 5148

The densest distribution of mills in the South is along the line of the Southern railroad, in the district known as the Piedmont. Of this group Charlotte in North Carolina is the natural centre: roughly, half the spindles and half the looms in the Southern States would be included within a circle around Charlotte of a radius of about 100 m. Of the remainder a large proportion is scattered over a wide area.

Much interest has been excited by this newly created Lancashire of a new type, and much speculation as to the causes that account for it has been elicited. An informal commission of Lancashire spinners and manufacturers crossed the Atlantic to make inquiries in 1902 and investigations have been undertaken by other persons56, and much has been written on the subject. A general explanation can now be framed without much difficulty, as in all probability most of the relevant facts have been brought to light. First and foremost the general development of the cotton industry in the United States must be emphasized. The industry was unquestionably foredoomed to expansion at this time, and the only question was where the expansion should take place. It was plain that the growth might be so great as to present the appearance of a new industry created with new labour rather than an extension of an old industry. It was not altogether surprising, therefore, that the exploitation of a new field of labour was thought of. The labour market of the North was comparatively exhausted; in less developed parts of the country larger supplies of intrinsically good labour might be looked for at lower wages. Skill was not a matter of much moment, because in the North it would have been necessary to incorporate much labour without previous experience in the industry, the work was intended to be of the rough kind upon which manual skill is least important, and it was intended to repose reliance for economy upon machinery in the main. The choice of new fields meant at the outset the sacrifice of some of the economies of localization, but so large an expansion was looked for that projectors did not despair of creating fresh industrial localization of sufficient magnitude to produce such economies as are derived from it, which, it must be observed, are inconsiderable in America, and have declined relatively with falling cost of transport and the adoption, as regards machinery, of the principle of interchangeable parts. And at any rate a new local industry would have a slight advantage in supplying markets in proximity to it.

These were the main general considerations, and the scale was turned in favour of the new locality (a) by the advantage of nearer supplies of cotton, and (b) by the known presence of much half-occupied white labour in the vicinity of otherwise suitable sites close to the cotton-fields. It must be borne in mind that the whole calculation had not to be reared merely upon an intangible theoretical basis. Cotton mills already existed in the South, and comparisons of costs of production, as things were then, afforded some groundwork for judgment.

As regards the first of the two special advantages mentioned above, the saving in the cost of carriage of the raw material is not commonly held to be high. Transport to the cotton ports is so well organized and sea-carriage is so cheap that Lancashire’s distance from the source of her raw material is not a very appreciable handicap. A good deal of the cotton that must be used in some of the Southern mills cannot be supplied locally because it is not grown in the neighbourhood, and the requirements of these mills are met by transport arrangements which at present cost a sum not altogether out of relation to similar costs in the New England States and Lancashire. The percentages of freight charges on raw material in 1900 were $2.18 in Georgia, $1.59 in North Carolina, $1.17 in South Carolina, and the amazingly low figure of $1.20 in Massachusetts, but of course some part of the explanation is the somewhat higher quality of cotton on an average that is worked up in Massachusetts. For some years, however, the saving in labour has been a most important economy. Large supplies of half-occupied white labour existed in the Southern States among the families of small farmers who flocked South after the Civil War, and in the districts of the decayed hand industry in the mountains of Kentucky and North Carolina. For small money wages much of this labour could be attracted to the mills. Negroes do not work in the mills; the reason is said to be partly their own disinclination and partly that they are not very efficient at factory work. As outside labourers, however, they have afforded important aid at a very trifling cost, but the expense of outside labour to a mill is never an item of much weight. The halcyon days to employers, when keen workers could be had for low wages, are now said to be past. The demand for labour was considerable, and as time went on additional supplies could be enticed only with the offer of better pay. In 1904 it was reported that some mills were unable to get fully to work for want of hands even at the improved rates. Again the Southern operatives have been visited by emissaries from the operatives of the New England States, which explains partly the present aspect of the wages question. Mr Pidgin, in his official report to the Massachusetts Bureau of Labour Statistics, questions whether a saving in wages can be expected to continue, and points out that though wages have been low the average efficiency of the operatives has not been high. Some, indeed, were sent to gain experience in Northern mills in the hopes that on their return they would spread the tradition of working at high pressure. Mr Pidgin is at some pains to measure labour efficiency in the South and North as far as it is possible to do so, but no simple sets of figures will prove very much. The value of the product per operative in 1900 was $1200 in Massachusetts, $1010 in Georgia, $937 in North Carolina and $984 in South Carolina, but the value of the product per operative depends as much upon the fixed capital charge per operative as upon the latter’s efficiency. And the amount of machinery used per head is higher in the South than in the North. The percentage of operatives to machinery in Massachusetts being expressed as 100, that of Georgia was 53, that of North Carolina 43 and that of South Carolina 55 in 1900. These figures must be borne in mind when the average numbers employed in a mill in different States are being considered: in 1900 the averages were 565 for Massachusetts, 273 for Georgia, 171 for North Carolina and 378 for South Carolina. Measured by quantity of machinery the sizes of mills would stand in quite different relations. Hours of work in the South are bound to fall and the abuse of child labour, which had unquestionably crept in, may be expected to discontinue entirely. The factory conditions of children are better now than they were, but in some places they are still very bad. In Georgia no children under twelve are employed, but infants without fathers may begin work at ten years of age, and according to Mr Pidgin’s report, “it certainly seemed as though the intention was honoured more in the breach than in the observance, or that there must be many widows in the neighbourhood of the cotton mills.” In North and South Carolina the employment of children under twelve is illegal, but in these States also conditions are recognized under which it is possible to employ them earlier. According to figures relating to 1900 the dependence on child labour in the Southern States is very striking. The proportions engaged at different ages in the three chief cotton-manufacturing Southern States and Massachusetts are as follows:

Men, 16 Years and over. Women, 16 Years and over. Children under 16.

Massachusetts 48.98 44.59 6.43

Georgia 39.98 35.52 24.50

North Carolina 42.22 34.23 23.55

South Carolina 44.43 28.72 26.85

It might be said that children are more useful when the work is rough, but this argument can hardly be regarded as accounting altogether for the great discrepancy as between Massachusetts and the South. The work is much rougher in the South: in 1900 the counts spun respectively in Massachusetts, Georgia, North Carolina and South Carolina were 25.10, 14.37, 18.83, and 19.04, and on the showing of the American census of 1900 spinning was getting finer over the last decade of the 19th century.

As contributory to the influences already recorded as accounting for Southern success it has been hinted that in the North employers have been less ready to welcome the new machinery, though in comparison with European rivals they would seem at first to have acted rashly. However this may be, the South enjoyed the important advantage that its industry began just after a great technical advance had been made. When Northern mill-owners were anxiously deliberating about the destruction of good machinery merely because it was antiquated in design, the fortunate Southern mill-proprietor was getting to work with appliances up to date in every particular. It will be easier to balance comparative advantages as between North and South when undertakers in the newer district are confronted by problems concerning replacements and alterations. The rapidity of Southern growth need not astonish those who have watched the operations by which new mills are frequently set up in Lancashire and remember that the American business man is more daring than his British cousin. Company promotion in the great financial centres, payment for machinery and other plant in shares, or partially in shares, a general diffusion of risks and pledging of credit, would explain even more rapid growth of industries of even greater magnitude.

Broad generalizations are difficult to frame, hard to establish and liable to be misleading; some generalizations relating to the features of the American cotton industry taken as a whole the author is tempted to venture nevertheless. Character of the American Industry. The characteristics of labour have already been incidentally commented upon. We have also noticed that the bulk of the work done is of a rough and simple character. In spite of American nationalism and the prevalence of protective sentiments it is said that there is still a prejudice in the United States against home-made fine cotton goods.57 “The product of the American system is a cloth which is, on the whole, distinctly inferior in appearance, ’feel’ and finish to that produced by the Lancashire system. To equal a Lancashire cloth in these respects an American cloth must not only be made of better cotton, but must contain more of it—perhaps 5% more. To this rule of inferiority there are, it is needless to say, exceptions, notably some of the American drills made for the China market. But the American home market, which absorbs nearly the whole of the product of American looms, is less exacting in these matters than the markets in which Lancashire cloths are sold.”58 It follows that the average counts spun in the United States are lower than in England, though they have been rising somewhat. Another feature of American spinning as compared with English is the high proportion of ring-frames to mules. In New England between 1890 and 1900 mule-spindles advanced by 100,000 and ring-spindles by nearly 2,000,000: in the South mule-spindles increased only from 108,500 to 180,500, while to the ring-frames 2,700,000 were added. To the general rule Rhode Island is the sole exception; here mule-spindles have increased and ring-spindles decreased; but in Rhode Island much of the fine spinning—for instance that for hosiery—is congregated.59 One explanation of the preponderance of ring-spinning is to be found in the character of American fabrics. Again most of the operatives are not of a kind likely to acquire great excellence at mule-spinning. To the Americans we largely owe the ring-frame, because their encouragement helped it through the difficult period when its defects were serious, though it appears to have been discovered independently in both countries.