60 II. FUNDAMENTAL STRUCTURE OF PROTOPLASM 



tained by this procedure from cucumber, tomato, and egg-plant, while 

 from dandelion and poplar were formed precipitates, the greater part 

 of which was composed of minute particles without double refraction, 

 only a small part presenting the refraction ; and from the other plants 

 doubly refracting precipitates were never obtained. 



From both tomato and egg-plant were sometimes obtained typically 

 ■crystalline masses with a definite shape and a sheen more intense than 

 that of the precipitates without double refraction, whereas in some 

 other cases only sediments were yielded appearing as though the mi- 

 nute particles themselves were exhibiting birefringency, " the doubly 

 refracting particles appearing to be so small. 



The nitrogen content of such crystalline masses or particles was 

 found to be approximately 16 per cent, the usual N value for ordinary 

 proteins, showing that there was no room for the presence of lipids, 

 and in fact no lipids could be recovered from such particles. By 

 these facts the writer was convinced that the lipids in non-coagulated 

 elementary bodies suspended in a plant sap were expelled from the 

 bodies by the addition of ammonium sulphate and that consequently 

 protein molecules were enabled to combine directly with one another 

 in their expanded state to become doubly refracting. 



The absence of doubly refracting property in the homogeneous 

 masses produced by the mere addition of acetic acid may depend on 

 the presence of lipids which may allow the protein molecules to par- 

 tially shrink or contract in an irregular manner. The homogeneous 

 masses without double refraction were actually proved to contain large 

 amounts of lipids as does usual protoplasm. When once coagulated, 

 animal elementary bodies, unlike plant particles, may hardly solve, if 

 any, their coagulation. Thi& may account for their inability to form 

 larger homogeneous masses not to speak of crystals by the addition of 

 ammonium sulphate. 



If the protoplasm protein of plant cells affected by a virus can be 

 obtained in such a crystalline form in which the protein structure to 

 act as the virus is retained without being damaged by the elimination 

 of lipids, the virus may be said to have been isolated in a crystalline 

 form. The crystallinity is therefore by no means a characteristic of 

 protoplasm protein having virus action. Crystals were actually obtained 

 from healthy leaves as above described. 



Since protoplasm protein is to be endowed with a virus action 

 through the structural alteration by the virus, some differences may be 

 expected in the crystalline form between "healthy" and "pathologic" 

 crystals so that a crystal peculiar to a virus may be obtained from a 

 plant affected by the virus. Bawden and Pirie (26) were able to crys- 

 tallize tomato bushy stunt virus as rhombic dodecahedra, whereas we 



