154 III. THE EVOLUTION OF VIRUSES 



Mueller (56) has shown that the infectivity of a filtrate of Rous 

 tumour may vary from nil to such an effective dose as 0.001 ml., and 

 moreover, it has been ascertained that the more slowly growing Rous 

 tumours are less filtrable than the rapidly growing ones, and that 

 occasionally the growth seems to lose its filtrability entirely, although 

 still transmissible by cell inoculation (57). It has been found that 

 the tumour of older chickens are tend to be more filtrable than that 

 of 3^ounger ones. In addition to this effect of host age upon the 

 filtrability, the age of the tumour itself, and whether the growth had 

 been transferred by cell suspension or filtrate previously, prove to be 

 important factors. It has been shown that non-filtrable growth be- 

 comes filtrable by the treatment with X-ray. In the case of Shope 

 papilloma, although the virus is readily demonstrable with wild cot- 

 ton-tail rabbits, if the tumour is transplanted to domestic rabbits, the 

 demonstration of the virus becomes very difficult, usually impossible. 



Mcintosh (58) has reported that fowl tumours induced by tar can 

 be transmitted to normal birds by cell-free filtrates, though this could 

 not be ascertained by some other workers. On the other hand, it is 

 known that if a fowl with a chemically induced tumour which cannot 

 be transplanted by cell-free filtrates, is injected with Rous virus, then 

 filtrates of this tumour will give rise to Rous tumour in other fowls 

 (43), a fact which might suggest independent nature of the virus. 

 This fact, however, can be interpreted as indicating the transmission 

 of the strong structural pattern of Rous tumour to chemically induced 

 one through the protoplasm particles of the former, thereby chemically 

 induced tumour is endowed with the strong structure to be held in its 

 protoplasm particles. Chemically induced tumours seem generally to 

 have structures weaker than those of naturally occurring tumours, 

 which may be the cause of the difficult demonstrability of the virus. 

 Oberling and Guerin have proved filtrability of the tumour induced 

 by methyl-cholanthrene in legs of chickens, but active fiiltrates were 

 only obtained in the fifth and sixth passage (59). 



3. The Maintenance of Assimilase Action 

 in Protoplasm Fragments 



Considerable damage of the assimilase action of protoplasm seems 

 to follow even Avhen the protoplasm is only slightly disintegrated, not 

 so completely as to pass through porcelain filters. 



On studying the number of metastases appearing in the lungs 

 following the injection of tissue emulsion of a transplantable mouse 

 tumour, Zeldman (60) claimed that, although the number of emboli 

 produced is effected by such single factor as the initial size of the 



