94 LEONELL C. STRONG 



Tabidation 

 Class 1 70 negative : 5 reactions ±1.40 or G.17% ± 1.81 



Classes 2, 3, 4 631 negative : 9 reactions ±2.01 or 1.40% ± 0.32 



Difference 4.77% ± 1.84 



The difference is thus 2.59 times its probable error 



This difference approaches mathematical significance, although 

 a difference of three times the probable error is not obtained." 

 It is safe to maintain that the difference would be significant when 

 larger numbers are obtained. The important point to notice 

 is the fact that the very young mice produce a susceptibility 

 curve that is on the same side of the middle-group curve as was 

 found to be the case with the dBrB tumor. 



Although the numbers are rather small for class 5, there is a 

 significant increase of positive reactions. In the following 

 chart bearing on this point the exceptional dBrA tumor was 

 included to increase the number of observations, but the results 

 are the same even if this exceptional tumor is eliminated (fig. 10). 



Class 5 55 negative : 12 reactions ±2.12 or 17.91% ± 3.16 



Classes 2, 3, and 4 673 negative : 26 reactions ±3.36 or 3.71%, ± 0.49 



Difference 14.20% ± 3.20 



The difference is thus 4.43 times its probable error 



The first point of the class 5 curve is possibly too low. 



g. Age susceptibility (dBrA). Susceptibility to the dBrA tumor 

 for the different age groups is given in figure 7 (heavy solid fine, 

 dBrA). The curve is of the same type as that for the dBrB 

 tumor (fig. 7), although it is not so striking. (This was to be 

 expected, since the dBrA tumor produced fewer reaction masses 

 than did the dBrB tumor.) 



h. Composite age susceptibility. The dotted line on figure 7 

 represents the total observations for both the dBrA and dBrB 

 tumors. 



We are justified, therefore, in concluding that the suscepti- 

 bility at different ages to transplanted tissue is independent of 

 the tumor tissue employed. 



^ The odds against chance alone being a factor when a difference of three times 

 the probable error is obtained is only one to 22.26, while for 2.6 times the prob- 

 able error it is one to 11.58. 



