216 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1949 
tobacco mosaic virus were found to possess properties quite similar to, 
yet in every case distinctive from, those of purified preparations of the 
ordinary strain. Spectacular progress has been made in the establish- 
ment of the nature of the chemical changes which accompany the 
mutation of tobacco mosaic virus. The amino acid composition of 
purified preparations of eight strains of tobacco mosaic virus and of 
two types of influenza virus has been determined. The results ob- 
tained with the strains of tobacco mosaic virus indicate that the 
mutation of a virus can be accompanied by the elimination of one or 
more amino acids from the virus structure, by the introduction of one 
or more new amino acids into the virus structure, or by a change in the 
concentration of one or more amino acids present in the virus structure. 
This work has great significance for it has provided the first informa- 
tion regarding the nature of the structural changes which accompany 
mutation. Extension of this work may reveal the exact nature of the 
chemical differences between virulent and avirulent virus strains, and 
provide important information regarding the mutation process in 
higher organisms. 
Attempts have been made to change the structure of tobacco mosaic 
virus by means of known chemical reactions in vitro in an effort to se- 
cure chemically modified active virus. Although several types of chem- 
ical derivatives of this virus were produced and were found to possess 
full virus activity, the inoculation of such virus derivatives to normal 
Turkish tobacco plants always resulted in the production of ordinary 
tobacco mosaic virus. ‘The results indicated that the chemical deriva- 
tives were converted to ordinary virus following their introduction into 
the cells of the plant, or, more probably, that the infecting molecules 
may not necessarily function as exact patterns for reproduction. De- 
spite these results it still appears that it may be possible to make 
changes in vitro similar to those which occur in nature, and thus secure 
a heritable chemical modification. Obviously this is a field in which 
important new results can be anticipated. 
Following the isolation of tobacco mosaic virus in the form of a 
crystalline nucleoprotein having individual molecules or particles about 
15 by 280 my in size, studies were undertaken in several laboratories to 
determine if other viruses could be obtained in purified form, mainly by 
techniques involving high-speed centrifugation. Some of these puri- 
fied viruses are crystallizable nucleoproteins having either rodlike or 
spherical particles. Some are nucleoproteins which have, as yet, not 
been crystallized. Others are large particles consisting of nucleo- 
protein, lipid, and carbohydrate, and possessing, in some cases, a 
degree of morphological differentiation which resembles that of organ- 
isms. Still other viruses have, as yet, defied isolation and purification, 
possibly, in some cases, because of extreme instability. The viruses 
