1906.] Experimental Analysis of the Growth of Cancer. 205 



success has continued till a level was reached which other tumours fell to at 

 the first essay. The protocol of this experiment is given in full on p. 211. 



■ ' i i i 



B4-0 830 660 WO 880 DAIS 



Fig. 3. — Graphic record of further propagation of the majority of the tumours obtained 

 in an experiment (50 Z) in which 90 per cent, of the animals developed 

 tumours. The diminished percentage of success reaches a minimum at the 

 first transplantation in some (51 Y, 51 W) ; in others after two transplanta- 

 tions (51 U to 52 T and 51 T to 52 U) ; while in others a third diminution 

 occurs before a minimum is reached (51 U, 52 S, 53 R) ; and a fourth diminu- 

 tion occurs in the series 51 T, 52 R, 53 Q, 54 L. 



The description of the graphic record has so far been confined to 

 illustrating the sequence of events in the tumours in the line giving the 

 highest percentages of success, and their ultimate fate. The other experi- 

 ments recorded in figs. 1, 2, 4 and 5 must now be considered. The tumours 

 propagated concomitantly with those marking the steps in the ascending 

 curve appear to form the ends of " side-branches " on that curve as a main 

 stem. Some of these " side-branches " also exhibited the upward tendency 

 and duplicated the behaviour of the main stem in that they also rose to a 

 maximum followed by a fall. The others appear to form descending side- 

 branches and to anticipate the ultimate fate of the ascending curve. If the 

 tumours on such descending side-branches be transplanted they may either 

 give a further fall, completely negative results, or gradually increasing 

 percentages of success till they in turn present a maximum followed by a 

 fall. 



The most careful attempts to maintain the percentage of success at a high 

 level in the direct line of descent therefore show that the condition leading 



