Differences in Transplantable Tumors. 47 



Beginning with the second week after inoculation, weekly 

 observations were made by means of palpation. If there were 

 any indications of the tumor, the relative size of the mass was 

 charted, as accurately as possible, on coordinate paper. Six 

 observations were made for each inoculation; it having been 

 shown by previous experiments that if any indication of growth 

 was going to develop it would appear in that period. 



Not one mouse in a total of 160 employed for this report ever 

 grew either tumor progressively for the entire period although 

 several indications appeared as shown in the chart (Fig. 1). 



A5 



1 



I 



h&eh Obstrved. 



Percentage Indications. 

 Weeks Observed. 

 Fig. 1. 



It will be noted that the dBrB tumor gave the greater percentage 

 indications throughout the experiment. 



According to the binomial theorem, the possibility of all six 

 points of the one curve being above the corresponding six points 

 of the other is as 63 : 1, thus indicating that the two curves are 

 significantly different, any odds above 27:1 being mathematically 

 significant. There is a mathematically significant difference be- 

 tween the per cent, of indications of the two tumors when the 

 results are massed, this being 8.6 times its probable error. The 

 analysis of the observations by successive weeks shows that the 



