312 REPORTS ON THE STATE OF SCIENCE. — 1915. 



The only other possible cause that could by itself explain the rise 

 in the accidents at any rate during the morning and afternoon is the 

 drinking of alcohol before starting the spell. This explanation has 

 been advanced by the Temperance Scientific Federation of Boston, and 

 taken up by certain employers. To prove this contention, however, 

 it would have to be shown firstly that the most debilitating effect of 

 alcohol on control occurs just about four hours after its drinking, and 

 not earlier or later, and secondly that such alcohol drinking is a regular 

 habit among the workers. 



The first point is far from established either scientifically or from 

 everyday — but not necessarily personal — experience. All that we can 

 say for certain is that if alcohol is taken at all in large quantities, the 

 attention and muscular control that avoid accidents are lost immedi- 

 ately, and in the first hour. 



The second point can certainly not be established at all in some 

 of our records. The women cotton-spinners (of Table IX. a), the 

 picked men workers at Hans Eenolds', at the Motor Companies, at the 

 National Cash Eegister Company, and the girls at Jacob's and 

 Cadbury's and the Denison Manufacturing Company (of Tables XVIII. 

 and. following), are certainly not all drinkers, yet all of them show the 

 same accident ' curve ' as other and possibly hard-drinking employees. 



If it be only fatigue, then, that can explain the middle hours, 

 what of the first hour of output and the fifth hour of accident immunity 

 that are left over? Here the explanation must be different in each 

 case, and such a difference may well be, since, as it was pointed out 

 in Section III. p. 299, variations in accidents and output are not 

 always measuring the same psycho-physical activities. It is now con- 

 tended that the small output in the first hour is due to ' practice ' and 

 that the fair accident immunity in the fifth or last hour is due to 

 anticipatory ' excitement ' (see Section IV.), both pulling in 

 an opposite direction to fatigue, and here more than overcoming the 

 fatigue effects; that this excitement does not affect the output at the 

 end of the spell, and that, contrary to all expectation, this practice 

 does not affect the accidents at the beginning of the spell." This con- 

 tention, founded on the facts in the tables below, is backed by the 

 somewhat theoretical suggestions of Section III. There it was ad- 

 vanced that the main psycho-physical activity directly measured by 

 output-rate variations was the changing of speed, and, by "Weber's 

 very definition, practice is a removal of pressure from the central 

 nervous organ manifested in an increase of facility, rapidity, certainty, 

 and regularity; but, further, that the main psycho-physical activities 

 specially measured by accident-rate variations were attention and 

 muscular control, and on these activities of the central nervous system 

 a ' psychomotor state ' of excitement would presumably be far more 



5 If, as was suggested in Section II., a curve of errors or mistakes would run a 

 similar course to the accident curve, it is interesting to note the conclusion of Binet 

 et Henri in 'La Fatigue Intellectuelle ' that for adding and memorising the second 

 hour of work, though faster, is less accurate than the first. Practice seems to 

 decrease errors no more than accidents. Pieraocini's curve of errors (Index D5) 

 confirms this in the afternoon but not in the morning. 



