SMITH— MECHANISM OF OVERGROWTH IN PLANTS. 443 



normally contains multipotent or totipotent cells which usually 

 remain dormant, but which under a strong stimulus are capable of 

 developing into either the whole organism or into some considerable 

 part of it, what is developed out of them depending on the degree 

 of differentiation of the cells at the time they are stimulated. We 

 may regard these leafy shoots (produced sometimes in great num- 

 bers where no buds occur normally) either as going to show that 

 potentially there is no dift"erence between germ-cells and young 

 somatic cells, or else that dormant "germ-cells" are widely and 

 abundantly distributed among the somatic cells, ready to develop 

 into the whole or a considerable part of the organism whenever a 

 sufficient stimulus is applied. Those who wish further details re- 

 specting these recently produced and peculiar crown galls contain- 

 ing fragments of the embryo plant are referred to a special paper 

 on the subject in the " Bulletin of the Johns Hopkins Hospital " for 

 September, 191 7. 



V. Bearings of These Discoveries. 



That these discoveries have many interesting bearings goes with- 

 out argument. Some of these bearings may be mentioned : 



(a) On the origin of insect, nematode and fungous galls ; 



(b) On the formation of thyloses in vessels; 



(c) On the origin, through absorbed poisons, of certain plant 

 diseases whose etiology is very obscure, such as peach yellows, 

 peach rosette, and the various mosaic diseases ; 



(d) On the origin, in the same way, of various plant and animal 

 monstrosities ; 



(c) On various problems of modification by slight changes in 

 environment ; 



(/) On possibility of normal wide distribution of dormant 

 germ-cells among somatic cells ; 



(g) And, finally, on the etiology of various human and animal 

 tumors. 



VI. Earlier Work and Reasons Why it Remained Sterile. 



I must here refer to some earlier work which remained sterile 

 so far as any influence on tumor etiolog)' is concerned (a) because 



