PLAN OF THE INVESTIGATION. 21 



thus be regarded as beginning with a " normal of the day," 1 which was 

 followed either by normal or by alcohol experiments according to a 

 predetermined plan. 



The non-alcohol periods and days are frequently called control 

 periods and control days in the literature. The term is misleading. 

 It would seem to imply that such experimental periods were occasional 

 and incidental to the main course of the experiments. In fact, the 

 non-alcohol experiment is as essential to the logical theory as the 

 alcohol experiment. Strictly speaking, the non-alcohol experiments 

 are not supposed to furnish controls of the validity of the other experi- 

 ments; they are supposed to furnish norms or base-lines from which 

 the alcohol experiments may or may not show characteristic differences. 

 In careful terminology, then, our non-alcohol experiments are not 

 controls, but basal or normal experiments, as Warren 1 and Rivers 2 

 properly call them. 



The normal or basal experiments were necessarily placed somewhat 

 differently in the various series, as they were arranged for the different 

 groups of subjects. The general arrangement for the main group was 

 as follows : For each series of tests, one normal session of three consecu- 

 tive hours preceded experiments with alcohol. Then followed one 

 session each with the smaller and larger dose of alcohol respectively, 

 given after the normal of the day. A final normal session concluded 

 the work of each subject in each series of experiments. The psycho- 

 pathic subjects began each of their two series of experiments with a 

 normal session. This was followed by an alcohol session for the same 

 series. On the fifth day a normal session was given for the combined 

 series. In the 12-hour experiments only two subjects were used, and 

 they were already familiar with the tests. The first day for each of 

 them was normal. On the second day hourly doses of 12 c.c. absolute 

 alcohol were administered after the normal of the day. In all cases the 

 alcohol was administered after dilution with 5 volumes of water, cereal 

 coffee, or other flavoring liquid. The arrangement as outlined above 

 provided for normal sessions before and after the alcohol sessions. 

 This gave an adequate normal base-line for the experiments, and 

 provided that any effects of practice in the various tests which might 

 appear in the alcohol sessions must also appear in the normal. 



'A term first used in alcohol experiments by Prof. J. W. Warren. (Journ. Physiol., 18S7, 8, p. 311.) 

 2 Rivers, The Influence of Alcohol and Other Drugs on Fatigue, London, 190S. 



