FRAENKEL-CONRAT AND WILLIAMS 



specific instance the protein was 

 sprayed upon the specimen screens at 

 a concentration of 0.1 per cent and the 

 RNA at 0.007 per cent. These are ten- 

 fold and sevenfold greater concentra- 

 tions of the two components than the 

 highest levels used in the assays of the 

 reconstituted virus (100 \ig./m\.). The 

 protein preparation showed no rods of 

 a length even as great as 100 niji in six 

 droplet patterns each of which had a 

 volume of approximately 3 X 10"'" ml. 

 The RNA material also showed no 

 rods in six droplets of approximately 

 this average volume. From these fig- 



Fig. 2. Electron micrograph of the 

 TMV protein used in the reconstitution 

 experiments. The particles are character- 

 istically disk-shaped, with central holes. 

 XI 20,000. 



ures it can be calculated that the con- 

 stituent protein and RNA solutions 

 contribute fewer than 5 X 10'^ typical 



269 



TMV^ particles per milliliter to the 

 reconstituted virus in the highest con- 

 centration assayed. Additional counts 



Fig. 3. Particles of the reconstituted to- 

 bacco mosaic virus. Their morphology 

 is identical with that of normal TMV, 

 except for a greater proportion of short 

 particles. The longest rod in this field is 

 about 300 m^ long. X60,000. 



made on the standard TMV used as 

 infectivity control showed that ap- 

 proximately 7X10** particles per milli- 

 liter, of typical length, are required to 

 produce the 10 lesions per leaf custom- 

 arily obtained. Inasmuch as several 

 preparations of the reconstituted virus 

 gave about twice that number of 

 lesions at the 100 |Ag./ml. level, we can 

 conclude that the starting materials 

 failed by at least a factor of 30 to con- 

 tain enough contaminating TMV^ par- 

 ticles to account for the final infec- 

 tivity. This conclusion rests upon the 



