There was some question about the oral route not being entirely oral because of injury to the esophagus during intubation. We have obviated that by allowing the animal to drink the brain extract or eat the whole brain without intubation, and, though the results are not quite as good as when the animal is intubated, they are quite striking, as Dr. Moloney has mentioned.

Perhaps the most interesting observation about the problem of the antigenicity of the agent is that even though the in vitro tests are not very striking, the neutralizing effect of the antibody is quite marked in in vivo experiments. For example, with our agent there is definite histologic evidence of leukemia 72 hours after the animals are inoculated with the virus. But if we start treating the animals at 72 hours with an antiserum prepared in rabbits, an antiserum that does not have a very high titer in vitro, the leukemic process may be reversed in a significant number of animals. If we allow these animals to go as long as a week after they have been inoculated with the virus, a time when they have gross evidence of leukemia, and treat them with the antiserum we are able to prolong their life about threefold. It would appear that the in vivo and in vitro effects are quite different as far as their presently measurable antigenic properties are concerned.

And finally, if I may anticipate Dr. Moloney's answer to Dr. Friend's question about hyaluronidase: In the passage of human leukemia to adult animals, we have used a technique not dissimilar to that used by Dr. Moloney, except that we were not using hyaluronidase in the preparation of our material. We found that if we used concentrated virus from the human brain we were able to induce leukemia in about 15 percent of adult animals of a resistant strain. But if we added hyaluronidase to this material, we were able almost to triple the incidence of leukemia and were able to get an approximately 50 percent incidence in a resistant strain of adult mice with material obtained from leukemic human brain.

Dr. Bryan: Are there any more comments on the specific phenomenon of neutralization?

Dr. Sabin (The Children's Hospital Research Foundation, Cincinnati, Ohio): I would like to ask Dr. Moloney whether we can assume that serums from mice and from rats that ultimately develop the tumors have not exhibited a neutralizing effect in vitro.

Also, is a normal serum an appropriate control for such a neutralization test? Is it possible-perhaps Dr. Moloney already has the data-that rabbits inoculated with normal mouse spleen, in which no agent is demonstrated, may have a similar effect if for no other reason than by aggregation of the particles in this particular extract. If so, would absorption of the antiserum that he has prepared eliminate this neutralising activity?

I have only one other point to make: When there are such difficulties in demonstrating antibody, which may be present in very low concentration, it may be worthwhile to try a procedure that virologists used years ago with the varicella virus before tissue-culture techniques were available.

I particularly recall the work of Russell Amies who was able to show that when highly centrifuged concentrates of material containing the virus were mixed with serum of a patient it was possible to demonstrate specific agglutination of the particles.

Four years ago, at a meeting in the Soviet Union, I heard a report-and I regret that I do not recall the name of the man who made it-in which it was claimed that ultracentrifuged sediments of various tumors could be agglutinated by antiserum prepared against them and absorbed with normal tissue, and this was shown in the electron microscope. The particles were very small. He was attacked by others in the audience for his conclusion that this showed that these particles were virus particles.

They said it is possible they could be antigenic cell particles. I bring it up here as one of the techniques that would be interesting to explore, particularly with the serum of the tumor-bearing mice and rats described by Dr. Moloney.

Dr. Moloney: In answer to Dr. Friend's question about hyaluronidase, Dr. Schwartz relieved me of that. We use it for essentially the same reason.

On titers, we have a 100 percent leukemia incidence in the 10-4 dilution. I do not know what the 50 percent point is. What is the relationship of the virus to Sarcoma 37? We think of it simply as being a "passenger virus" in Sarcoma 37. Since 1908 this tumor has been in many laboratories. Apparently the Sarcoma 37 serves as an ideal medium for the maintenance of the leukemia virus, though not particularly for its multiplication. There is, however, multiplication to some extent. We have not been able to induce Sarcoma 37 with the leukemia virus, but only these lymphocytic neoplasms.

Graffi has expressed a similar opinion on the relationship of the myeloid leukemia virus to the many sarcomas and carcinomas from which he recovered it.

In answer to Dr. Sabin, our completed serological studies are still preliminary and not ready for publication; we have that work in progress. The reason why we did not publish the preliminary results was that normal serum is a very poor control, as Dr. Sabin has emphasized. We now have in experiment absorbed serum and serum produced against normal spleen extracts. The antiserum described here was produced against leukemic spleen extracts. We appreciate your last suggestion concerning our work.

No serological tests have been done on the rats.