44 Effect of Alcohol on Psycho-Physiological Functions. 



adjustment of the apparatus to the particular individual and thus to 

 favor the production of comparable results. 



The mean variation within any one series of consecutive reflexes is 

 usually for the latent time not larger than 1.5 <r, which is from 3 to 5 

 per cent of the average latency. The mean variation for the ampli- 

 tude is from 1.5 to 4.0 mm. for first reflexes. Here the two intensities 

 of stimulation must be considered. Figure 1 is composed of parts of 

 three records. The amplitude of the reflexes from 30-gram blows is, in 

 general, about one-half that resulting from 50-gram stimuli. The 

 amplitude with the 50-gram hammer is very large at the beginning, then 

 gradually decreases, although the curves commonly cross and recross 

 each other. They show greater variation among themselves than is 

 found with reflexes from 30-gram stimuli. Referring to tables 3 and 4, 

 section i, we find that the usual mean variation in millimeters for ampli- 

 tude of reflexes with 50-gram stimuli is approximately 50 per cent 

 larger than that with 30-gram stimuli. However, if the mean variation 

 is expressed in per cent, it is found to be larger for the 30-gram stimuli, 

 the average with the first reflexes being about 30 per cent and for 50 

 grams nearer 20 per cent. The amplitude of the patellar reflex is there- 

 fore a much more variable and, it may be, unreliable quantity than 

 the latency is. The mean variations for latency are practically iden- 

 tical with both intensities of stimulation, notwithstanding that the 

 more intense stimulus produces reflexes with larger variations in am- 

 plitude, which variations should, according to Hoffman, 1 react on the 

 latency to some extent. From figure 1 it can be shown that the error 

 in reading the curve for latent time will be larger in the case of those 

 reflexes produced by the 30-gram stimulus. In the first place, the 

 depression at the extreme left of the curve, the beginning of which fixes 

 the moment of stimulation according to the technique employed, is less 

 well marked for the 30-gram stimulus, and the curve at the moment of 

 muscle thickening leaves the base-fine more gradually. The errors in 

 reading these curves will tend to cancel each other, as far as the average 

 latency is concerned, but will, on the other hand, increase somewhat 

 the mean variation. As both points are more definitely fixed for meas- 

 uring the curves produced by 50-gram stimuli, the errors of observa- 

 tion should be smaller, and as the mean variation for latency in the case 

 of both reflexes produced by both intensities is almost identical, it 

 seems that the actual variations in latency for the reflexes from the 50- 

 gram stimulus must theoretically be somewhat larger. Probably a 50- 

 gram stimulation was too strong for the best condition of normal relax- 

 ation with this subject, as the large contraction of the quadriceps muscle 

 of the left leg must almost necessarily have tended to produce a pre- 

 paratory stiffening of the whole body, even though the subject said 

 "Ha" after each blow, and the excessive contraction of the muscle is 



1 Hoffman, Arch. f. Physiol., 1910, p. 223. 



