1906.] Experimental Autolysis of the Growth of Cancer. 213 



represent the series of experiments giving a maximum at 48 E, the con- 

 stitution of the parent tumours as revealed by the percentage of fragments 

 developing into tumours can be depicted by subdividing each large square 



Fig. 6. — Histological differences between cells in a tumour apparently healthy and homo- 

 geneous to the naked eye. Islands of clear cells, whose nucleus and proto- 

 plasm have little affinity for stains, are surrounded by cells whose nucleus 

 and protoplasm stain intensely. The latter are more numerous on the surface 

 of the tumour alveoli. This differentiation is very frequent in tumours 

 undergoing spontaneous absorption. 



into 100 small squares each representing an implanted fragment, and 

 blackening as many as there were implantations which did not yield tumours. 

 In the accompanying diagram (fig. 7) the clear part of each large square 

 represents the percentage of success attending the transplantation of a tumour 

 arising from a single small square in the one before it. We may imagine that 

 the blackened part of each square (fig. 7) represents those implanted frag- 

 ments of tissue which, healthy at the time of inoculation, are already on the 

 way to degeneration and do so degenerate immediately after transplantation. 

 The continuous diminution in successive subinoculations which this black- 

 ened part undergoes, as the number of fragments developing into tumours 



