Patellar Reflex. 43 



ments. The comparison series of values for the alcohol days, +1.7, 

 + 1.0, +2.3, —1.0, and +0.5 a, show in the first three periods an in- 

 crease in the latency with a 30-gram stimulus which is greater than in the 

 similar periods under normal conditions. To secure a statement for 

 this difference in degree of increase of latency, which we assume to be 

 the effect of alcohol, the normal differences are subtracted from the 

 alcohol differences. The results thus obtained, shown in section in, 

 for the 30-gram stimulus, are +1.4, +0.3, +1.0, —2.5, —0.5 <r for 

 periods 2 to 6, respectively. These differences averaged, according 

 to the method of Dodge and Benedict, equal —0.6 a (see next to last 

 column, section in). The average latency for all of the first-period 



measurements, both normal and alcohol days (see section i, — 



= 34.5 a), forms the basis for computing the percentile effect of alcohol. 



+ 1.4 

 For example, the computation for period 2 will be: oa z. = +4.1 per 



cent for the 30-gram stimulus (see section in, period 2) . The percen- 

 tile values for the other periods and for the amplitudes are computed 

 in a similar fashion. Finally, the percentage effects on the reflexes 

 produced by the two intensities of stimulation are averaged. 



Before discussing the results certain features of the data should be 

 noted. The strength of stimulus changes the latency and amplitude of 

 the reflex. This has been shown by Exner, 1 Rosenthal, 2 Dodge 3 and 

 others, and can be seen clearly in figure 1, comparing reflexes with 30- 

 gram and 50-gram stimuli, and from table 3, section i, in which the 

 averages for 30-gram and 50-gram stimuli are given separately. The 

 difference in the relation of the two intensities of stimulation to the 

 second reflex, which came within the refractory phase, can not be 

 demonstrated in figure 1, but may be seen in table 4, section i. From 

 these values it is apparent that the average latency, L', is increased 

 with the 50-gram stimulus in contrast to the condition for L, 

 while A' is also usually larger with the 50-gram stimulation. The 

 variability of the patellar reflex is a feature to which Dodge and Bene- 

 dict call special attention, illustrating normal variations by the case of 

 their Subject II (see page 46 of the Dodge and Benedict report), whose 

 values for September 23, 1913, first period, were L, 35 a, A, 21.4 mm.; 

 I/, 42 a; A', 7 mm.; while for October 8, 1913, the similar values were 

 L, 51 <t; A, 2.4 mm. ; L', 50 a; A', 2.2 mm. The data which have been 

 presented in tables 3 and 4 show no such large variation. Probably 

 this was because the experimental days were successive and the con- 

 ditions more homogeneous and constant throughout the week. In 

 working with one subject it is possible, moreover, to maintain a better 



1 Exner, Archiv f. d. ges. Physiol, 1874, 8, p. 526; and Vdrtrag auf dem allg. fort. Intemat. Kon- 

 gress fur Elektromedizin u. Radialogie, Amsterdam, 1908. 



2 Rosenthal, Arch. f. Psychiatrie u. Nervenheilkunde, 1885, 15, p. 184. 



3 Dodge, Zeitschr. f. allg. Physiol., 1910, 12, p. 28. 



