84 R. MARKHAM 



is an excellent solvent for proteins, but the nucleic acid remains in tlie 

 aqueous phase. The extraction with phenol was repeated twice more, and 

 excess phenol was removed from the water by means of ether. The aqueous 

 solution was tested on Nicotiana glutinosa. The infectivity was then found 

 to be equivalent to 0.1 % of that of an equal amount of nucleic acid as it 

 existed in untreated virus. \Vhen one considers the relative lack of infectivity 

 of untreated virus, and that some 10^ particles of nucleic acid were therefore 

 needed to cause a single infection, a certain amount of faith is required in 

 the interpretation of the results, particularly if one considers that the absence 

 of active virus can only be controlled by indirect methods. 



In particular, the infectivity possessed by the nucleic acid preparations 

 is very labile to the action of pancreatic ribonuclease at concentrations of 

 the latter which have little effect on whole virus (Gierer, 1957). Conversely, 

 the activity is much less affected by antiserum to the whole virus, though 

 it is indeed surprising that the serum used did not contain enough ribo- 

 nuclease to inactivate the nucleic acid. 



A rather more striking piece of evidence in favor of the theory that the 

 nucleic acid is itself infectious is given by Schramm and Engler (1958), 

 who state that the course of infection with nucleic acid is different from 

 that when whole virus is used. Using the ability of the virus to produce 

 systemic infection in tobacco as a sensitive test for the production of virus, 

 they fomid that they were able to detect virus in tobacco plants infected 

 with the nucleic acid about 20 hours after inoculation. In plants inoculated 

 with whole virus this state was not attained until some 12 hours later. This 

 suggests that the nucleic acid, not having to emerge from its protein sheath, 

 is able to multiply without so long a time lag. 



Unfortunately, the figures which Schramm and Engler produce to sub- 

 stantiate this claim contain a number of improbable features. First, they 

 claim that they get 54 % of infections with solutions which only contain one 

 particle per miUilitre. Apart from the inherent improbability of this figure, 

 which is many orders of magnitude greater than normally obtained, they 

 give figures showing that only 4 plants out of 25 inoculated with virus a 

 million times as concentrated became infected. It seems likely that their 

 experiments contained a source of uncontrolled error. Leaving this aside, 

 however, there would seem to be no reason why one should query the fact 

 that detectable virus appears some 12 hours earlier if nucleic acid is used for 

 the inoculations rather than whole virus. A similar conclusion was reached 

 by Siegel et al. (1957) on the basis of measuring the ultraviolet light sensitivity 

 of tobacco mosaic virus and virus nucleic acid after inoculation into plants. 

 The nucleic acid used by these workers was prepared by treatment of the 

 virus with dodecyl sulfate, as was that made by Fraenkel-Conrat (1956), and 

 in both cases the low infectivity observed by Gierer and Schramm was 



