414 JAMES HILLIER 



remained for the electron microscope to provide accurate pictures 

 of the shapes of the particles. Since most of the smaller viruses are 

 spherical, the electron microscope was not able to contribute very 

 much beyond this verification of shape and size. In fact, the statis- 

 tical methods of working with these viruses were more accurate and 

 at the same time permitted the study of the viruses under more nor- 

 mal conditions. 



In later work, however, this situation changed considerably. 

 For instance, Wyckoff (12,13) and co-workers have studied the de- 

 tails of the crystallization of spherical virus particles. By evapo- 

 rating aqueous solutions of purified plant viruses, Wyckoff found that 

 the elementary particles were left in quite regular arrays over small 

 areas. The arrangement was the normal close packing that one 

 would expect from spherical particles. However, when he studied 

 replicas of the surfaces of small crystals of the elementary particles, 

 the crystals being formed from concentrated salt solutions, he found 

 that the arrangement was quite different. In fact, the same particles 

 now became arranged in a rectangular network and, moreover, the 

 individual particles were seen to be considerably distorted. Such 

 results could hardly have been obtained by any other method than 

 by the electron microscope and they have obvious extensions in the 

 study of the structure of other protein or macromolecular substances. 



Another problem that recently has been given considerable 

 attention is in connection Avith the length of the tobacco mosaic 

 virus particles. It has been known for some time that old purified 

 solutions of tobacco mosaic virus have particles that show a rather 

 wide range of lengths. On the other hand, the particles examined 

 in dried films of the juice of a hair cell from an infected leaf are 

 remarkably uniformly 280 m/i in length, a result which Oster and 

 Stanley (14-) consider conclusive evidence that the infective unit of 

 the tobacco mosaic disease is a particle 15 m/i wide and 280 m^t 

 long. To obtain the distribution of lengths for a large number of 

 particles in a given preparation is a direct and relatively simple matter 

 with the electron microscope. To obtain similar measurements in- 

 directly and by statistical methods is, of course, exceedingly difficult 

 if not impossible. 



In the virus problem some of the most interesting work has been 

 done in connection with the infection of bacteria by the bacteriophage. 

 Here the electron microscopist has been fortunate because many of 

 the known bacteriophages have a very characteristic morphology 



