90 LEONELL C. STRONG 



that reason the difference between corresponding points at 

 the latter end of the curve would not be as great as between 

 points at the other end of the curve. The slight rise in the 

 dBrA curve is not real, but due to the dying off of several nega- 

 tive individuals during an epidemic in the summer of 1920. 



1. Total dBrB 876 negative : 80 reactions ±5.74 or 9.23% ± 0.67 



2. Normal dBrA 705 negative : 15 reactions ±2.58 or 2.08% ± 0.36 



Difference 7.15% ± 0.76 



The difference is thus 9.40 times its probable error 



d. Age at inoculation (dBrB). All mice were separated into 

 five groups according to age. The five classes arbitrarily se- 

 lected were as follows: class 1, to 3 weeks old; class 2, 4 to 

 7 weeks old; class 3, 7 weeks to three months; class 4, adults, 

 and class 5, old mice. The mice were so classified so as to 

 include groups to represent the approximate life-cycle of a 

 mammal (youth, puberty, adolescence, maturity, and senes- 

 cence). The dBrB tumor may be considered first and the 

 data for the three middle classes (2, 3, and 4) compared. 



The data may be summarized: 



Class 2 87 negative : 10 reactions d=2.02 or 10.30% ± 2.09 

 Class 3 146 negative : 10 reactions ±2.06 or 6.41% ± 1.32 

 Class 4 523 negative : 43 reactions ±4.25 or 7.59% ± 0.76 

 The difference between the percentages are: 



2-3 3.89% ± 2.47 or 1.57 times probable error 

 2-4 2.71% ± 2.22 or 1.22 times probable error 

 3-4 1.18% ± 1.52 or 0.77 times probable error 



The three classes are remarkably similar, there being no sig- 

 nificant difference between any two classes. We may therefore 

 group these three classes into one general class, so as to simplify 

 the comparison with class 1 and with class 5. 



Individuals from class 1 (0 to 3 weeks old) gave a very high 

 initial percentage of palpable reactions. Unfortunately, the 

 positive individuals did not survive the whole length of the 

 experiment, so that the class 1 curve drops to zero with the 

 seventh week. We have made the curve end with the sixth 

 week becavise we feel that that is the only way to represent the 

 true state of affairs. Very young wild mice, if they showed 



