men drink more alcohol; at all events, their convictions for drunkenness are three or four times more numerous

than those of women. In correspondence with the drinking habits mentioned, it was found that, if the average number of accidents per hour in the men during i the first spell were taken as 100, it fell to 75 in the is second spell, and to 62 in the third spell. In the women l the relative number of accidents in the three spells varied as 100, 89, and 72, respectively. As already stated, t this diminution of accidents during the course of the night was due largely to the increasing calm of the night-shift workers; and such recovery of mental equilibrium could not have been due wholly to the dish appearance of the artificial excitement produced by alcohol consumption. That it was due partly to this artificial cause appears to be shown by the fact that, s as the sobriety of the workers increased, the number of accidents in the first spell grew relatively less and d less. In the men the proportions of accidents changed i even more than in the women.

In the day-shift workers there was no evidence of any alcoholic effect except at the time when they were working a twelve-hour day. The number of accidents was then at a maximum on Monday, and dwindled rapidly during the course of the week till it reached a minimum on Friday, some 30 per cent. smaller than the Monday maximum. On Saturday it suddenly shot up again and nearly reached the Monday maximum; and the entire disappearance of this week-end excess of accidents when the ten-hour day was substituted for the twelve-hour day seems to point to week-end drinking as its cause. Wages were paid on Friday evening, and it seems probable that the over-fatigued workers were inclined to indulge themselves then, and on the next two days. On most nights of the week they did not stop work until 8.30 p.m., and so had not much opportunity; but in the ten-hour day period they stopped work at 6 p.m., and so they presumably took their alcohol in more moderate quantities, and distributed it evenly throughout the week.

Assuming the substantial correctness of the conclusions so far arrived at touching the causes, do they

9

uggest any means for reducing the number of accidents ? an one do anything to diminish the inattention of the ight-shift workers when they first come to the factory, nd bring them into the, condition of mental calm into hich they pass in the small hours of the morning? t is even more important to bring the day shift to a imilar mental state. Thus it was found that, in all the actories, the accidents were more numerous in the day hift than in the night shift, the average excess being per cent. This excess of accidents occurred in spite f the fact that the alcohol factor operates more powerally on the night shift than on the day shift. It is robable that the difference is due chiefly to the psychical ictor. Though the day-shift workers are in a dull ental state when they first come on to work, yet for ost of their working hours they are in a greater mental orment than the night shift. Thoughts of home life ad its troubles, of friends and relations, of pleasures nd amusements, all take attention from the work in and. The simplest way to increase the mental con›ntration of the workers would be to forbid all conversaon in the shops, except that relating to the work in and; for even a minute's talk with a friend on matters Fot related to the work may initiate a vigorous chain of hought, and much mental unrest and distraction. The ardship, if hardship it be considered, could be mitigated y reducing the hours of work, and by the provision of omfortable canteens, rest-rooms and reading-rooms in he factory. In other words, let the employees conntrate all their energies on their work during shortened Working hours, but let them have every comfort and onvenience possible during non-working hours.

The system of posting 'safety bulletins' and other formation relating to accident prevention at the trance to each of the workshops is certainly worth ying. As it has proved useful in the United States might be expected to do some good in this country, ough the innate conservatism of our average working an is difficult to combat. It would be essential that he bulletins should be changed at frequent intervals,

is done in the States, for otherwise the workmen ould glance at them once, and ignore them and their ggestions for ever after. One knows that, of men

engaged on the same kind of work, certain individuals frequently suffer accidents, whereas others almost invariably avoid them; in other words, some men are inherently clumsy or careless, while others are the reverse. Suitable education may at least teach the inherently clumsy man to exercise more care.

The harmful effects of consuming alcoholic liquids can be combated in several ways. In the day shift, the temptation for the workers to drink during the mid-day meal hour could be diminished by offering them induce ments to take their meals in the factory. The number of workers catered for in munition-works canteens increased rapidly during the war; in 1917 it amounted to 800,000, in 1918 to nearly a million.* The canteen system is extending to other factories; and the more general it becomes the better will it be both for employers and employed. It would be quite worth while to provide food at slightly under cost price, as the expense would be more than repaid by the increased vigour of the workers. The workers would be able to rest for the whole of their break, instead of having to hurry to and from home; they would get better-cooked food; and they would be relieved from the temptation to get a drink. Moderate drinking in the evening, when work is finished, would have little or no influence on the chance of accidents, or on the quality and quantity of the output, as its effects would have passed off before work began next morning. The temptation for night shifts to have a drink just before coming on to work can be reduced by arranging that the shifts start shortly before the time at which the publichouses are opened. This was usually done at all the factories investigated; but in two of them, for certain periods, work did not start till after opening time, and the relative number of accidents in the first spell of night work increased, presumably in consequence.

The preventive measures thus far suggested are not very striking or original, and the sceptic may consider their probable effectiveness to be very small. What it would be in actual fact one cannot tell without direct

* Fourth Report of Central Control Board (Liquor Traffic), 1918, p. 11. [Cd. 9055.]

/

xperiment. But certain other preventives will now be liscussed, the efficacy of which is demonstrable by actual igures. The first of these is temperature. Every one nows that in very cold weather the unprotected hands re liable to get numb; and such numbness means loss f manual dexterity or increased liability to accidents. One would imagine that the effects of cold would be pecially great in engineering shops, where the hands re constantly in contact with cold metal tools and other netal objects, and are frequently wetted the whole day hrough by the stream of soapy water in which many aetal articles are turned. Yet, curiously enough, no irect observations appear to have been made hitherto pon the relationship between temperature and liability o accidents.

In order to acquire such information, I determined he temperature variations at the fuse factory coninuously for six months by means of a recording hermograph. The instrument was placed in the centre f one of the large shops (200 × 200 feet), and, by comarisons with thermometers placed in other shops, it vas found to give a fairly accurate idea of the mean emperature of most of the factory. On an average, ccidents were at a minimum when the factory temperaure was about 67° F., though they were only slightly aore frequent so long as it lay between 60° and 70°. At emperatures above these limits they began to increase aore rapidly in number, and at 77° were 30 per cent. nore numerous. At lower temperatures they likewise acreased, being 18 per cent. more numerous at 56°; but he artificial heating of the factory was so good that had no opportunity of observing the effects of real cold. was, however, able to make observations of this kind t the shell factories, which were situated in a more ortherly and bleaker district than the fuse factory.

In these factories the heating was only moderate; nd, as large doors had frequently to be opened to dmit trucks of shells, the temperature in parts of he shops temporarily fell almost to that outside. I id not record the temperature in the factories themelves, but obtained complete records of the external emperatures; and it was instructive to see how the roportion of accidents rose steadily on passing from

the autumn to the winter months, and fell again with the best advent of spring. It was found that at one factory the evid women experienced on an average nearly two and a half abso times more accidents when the external temperature Othe was at or below freezing-point than when it was above king 48°, while the men experienced twice as many accidents hous At intermediate temperatures the accidents were inter in mediate in number, or there was a clear and close plie relationship between accident frequency and temperais ture. Hence, there can be no doubt that low temperature stan is one of the most important causes of accidents. More over, it is a cause which can be and ought to be avoided as in practically all indoor industries. Glaring instances of ed its occurrence are often noted by factory inspectors. We ai read* of a clothing factory with a temperature of 39°, of p a large provender mill with one of 28°, and of the carding er room at a flax mill with one of 35.5°. Abnormally highl temperatures are likewise not infrequent. Temperatures e of 109° and 113° are recorded in cotton mills, and of 106 e and 111° in print works.

Another well-known cause of accidents which can in easily be remedied is defective artificial lighting. The r Departmental Committee on Lighting in Factories and o Workshops, in the weighty report issued by them inc 1915, adduced indirect evidence to show that inadequate lighting is a contributory cause of accidents; and their conclusions were confirmed by the evidence of witnesses regarding accidents in foundries, shipbuilding yards and cotton mills. I myself did not obtain very striking data, as the factories in which I worked were all well lit. The m only type of accident which was considerably affected by artificial light was that of foreign bodies in the th Apparently the workers bend more over their work when the lighting is artificial, and so get more metal turnings and particles of emery jumping into their eyes Though, as already stated, night-shift accidents, taken as a whole, were distinctly less frequent than day-shift accidents, eye accidents were 30 to 60 per cent. more numerous in the worst-lit factory. At two of the other factories they showed a smaller excess than this, but at

eye.

* Annual Report of Chief Inspector of Factories and Workshops, 1912, pp. 81 and 124.