PHYSICAL AND CHEMICAL PROPERTIES OF VIRUSES J 51 
sub-units are not, of course, infectious, and the nucleic acid is split 
off most of them. 
When completely dry, the gel is horny, yellowish in colour, and 
only feebly birefringent. Wyckoff has shown that under the electron 
microscope the wet gel does not appear to be the hexagonally ordered 
array, postulated by Bernal and Fankuchen, but consists instead of a 
flat sheet, or several layers of sheets (Plate XV). 
Although the tobacco mosaic virus, when purified, does not form 
three-dimensional crystals, infected plants do contain true crystals 
in the form of hexagonal plates, first described by Iwanowsky and 
called by him “striate material.”” When these plates are touched in 
the cell with a micro-manipulator, they break up into the fine needles 
first described by Stanley, and it is thought that they are a complex 
of virus and some other component. According to Kausche 
crystals of this type can be produced by the action of plant sap upon 
the ammonium sulphate precipitate of the virus, but no one has 
confirmed this rather unlikely statement. 
The tobacco mosaic virus is stable over a range of pH 1-5-8-5 but 
is not stable to alkali; it precipitates at its isoelectric point, pH 3-4. 
The virus consists entirely of protein and nucleic acid, the latter 
being about 5-6 per cent of the total. 
It is not possible to separate the nucleic acid without denaturing 
the protein and thus destroying the activity of the virus. Any treat- 
ment which denatures the protein separates off the nucleic acid, and 
there is a great deal of evidence which suggests that the nucleic acid 
is essential for multiplication of the virus. 
The nucleic acid is of the ribose type and in 1947 Schwerdt and 
Loring isolated the brucine salts of uridylic acid and guanylic acid 
from hydrolysed nucleic acid. The elementary composition of 
nucleic acid is as follows: C 35 per cent, H 42 per cent, N 15 per cent, 
P g per cent, Purine N 8-5 per cent, CH,O 31 per cent. There seems, 
however, to be too little carbohydrate. 
The protein component of tobacco mosaic virus is unusual; instead 
of being the basic protein of the histone or protamine type which 
one might expect in a nucleo-protein, the protein component is 
acidic, with an isoelectric point about pH 5:3. The general com- 
position resembles that of casein (Chandler et al., 1947), and has no 
preponderance of basic amino acids. 
There are sixteen amino acids in all (Knight, 1947), the major one 
being arginine 9-8 per cent, the others are lysine 1-47 per cent, aspartic 
