III. IMPERFECT VIRUSES 153 



the writer's concept of the nature of viruses it is never an unex- 

 pected result. In addition to this tumour, papilloma vi^hich affects cot- 

 tontail rabbits is known to be transmitted by a virus, and which is 

 commonly called Shope papilloma for it was described first by Shope 

 as the fowl tumour is called Rous tumour after its first observer. 

 Shope papilloma is characterized by the presence of warty or horn-like 

 protuberances on various parts of the rabbit body, whereas human 

 wart was also ascertained to be infectious through a cell-free filtrate 

 long before the virus nature of Shope papilloma was confirmed. 

 Mammary cancer of mice is likewise believed to be caused by a virus. 

 This virus has been demonstrated in the milk and also in high con- 

 centrations in both the lactating mammae and mammary cancer of in- 

 fected mice. The injection of so little as 10"^ to 10"^ gm. equivalents 

 of lactating breast tissue has been sufficient to infect suitable host 

 animals (51). The virus nature has been confirmed also with a kidney 

 cancer of a frog, Rana pipiens. 



Again, fowl leucaemia can be regarded as a kind of malignant 

 tumours, and it was found that the infection could be transmittted by 

 0.000,0001 ml. of cell-free plasma, and that the virus could withstand 

 drying for at least 54 days (52). Furthermore, both a myxoma and 

 fibroma of rabbits can be transferred by filtrates of the tumour sus- 

 pensions, and it is said that the virus of one of these diseases has a 

 remarkable relationship to the virus of the other, and in view of the 

 pathological changes caused by the viruses, these two diseases are 

 considered to lie between tumours and usual virus diseases (53). 



In these two tumours and also in fowl leucosis natural infections 

 are said to be possible. Wound-tumour diseases of plants is likewise 

 a tumour in which a virus is involved ; this tumour is transmitted 

 by an insect as in the case of usual plant-virus diseases (54). Never- 

 theless, in many other tumour diseases natural infections may fail to 

 occur, although the transmission can sometimes be brought about arti- 

 ficially through cell-free filtrates ; for instance, with Rous tumour it 

 has been definitely confirmed that natural infection is entirely impos- 

 sible despite the fact that a virus is involved in the tumour. 



Carr (55) has emphasized the fact that the greatest recovery of 

 infective virus from Rous tumour represented a yield of only one in- 

 fective particle per 20 tumour cells. This appears to be surprising, 

 since even when peculiar cytoplasmic particles only, containing large 

 amounts of nucleic acid, could act as the virus, at least scores or 

 hundreds of virus particles would be obtained per cell. This might 

 be attributed to the requirement of a vast number of particles to 

 cause a tumour because of the weakness or the lability of the tumour 

 structure. 



