The same type of phenomenon was observed in primary cultures. Visible, distinct colonies of changed cells, easy to count, have been observed in 2- to 3-month-old cultures. It has been interesting to note that in several cases in which the distribution of the primary cells was uneven, the major part, in some instances practically all, of the colonies of changed cells occurred in the parts of the culture in which the density of unchanged cells was lowest. This observation agrees with Rubin's observation of the Rous virus transformation system in tissue culture, and it could be explained likewise by an inhibiting effect of the unchanged cells upon the development of the changed cells.

In another type of experiment we have studied the effects of filtrates prepared from changed cells (FL cells). The cells were broken down and the cell material was passed through a series of Millipore filters. Primary amnion cultures inoculated with filtrates showed changes not observed in uninoculated cultures. Filtrates passed through the 450 or 300 mµ Millipore filter have, within 2 weeks after inoculation, caused macroscopically discernible plaque-type lesions in the cell sheets. The lesions were not exactly countable, but were similar in gross appearance to various plaque types we have seen in monolayers infected with animal viruses. The plaquelike lesions, however, were different from the type seen, for instance, in polio virus-infected cell sheets. When observed in the microscope, they were filled with rapidly growing, small, clear cells with a morphology different from that of cells outside the plaquelike areas.

We have one more experiment in which the effects of filtrates have been noted. Twelve amnion cultures were inoculated with Millipore filtrate, and 12 cultures were left uninoculated. Several months later changed cells appeared in 5 inoculated cultures, whereas nothing happened in any uninoculated culture.

I am not prepared to interpret these findings. There are still many control experiments which have to be carried out. I thought, however, that this observation was worth presenting at this seminar to show that interesting things can also happen to human cells.

In spite of great progress in virology, many points are still confusing, but once in a while the virus particles help us by getting into a controlled, orderly system-for example, when particles of polio virus arrange themselves in crystalline arrays in the cytoplasm of infected HeLa cells.

Dr. Dalton (National Institutes of Health): I want to acknowledge the beautiful demonstration of material shown by Dr. de Harven. I can agree with his remarks, in regard to the Moloney agent and its great similarity to the Friend virus, that in the beginning the particles are the A type and then apparently, on aging, become the C type. However, the size of the particle itself is considerably greater, closer to 100 mµ than 85 mµ.

In the cell of origin of the Moloney agent, presumably in Sarcoma 37 in which the particles are present intracellularly, the particles have a single outer membrane of 86 mµ diameter, while those in an extracellular position have a diameter closer to 100 mµ.

It should also be mentioned that Miss Mottram and I have been examining the material provided by Dr. Alice Moore, and in several lymphomas derived from different sources there were seen doughnut-shaped particles similar to those described in Dr. Friend's material.

Dr. Bonar: Will Dr. de Harven elaborate on the difference between the extracellular A type and the intracellular doughnut-shaped particles?

Dr. de Harven: The difference is especially obvious after lead hydroxide staining. After such treatment, the inner shell of the leukemia virus appears extremely dense, far more so than the outer shell. In the lymphoma particle there is not the same difference, and the ratio between the diameters of the outer and the inner viral membranes is also different.

Dr. Gey (Johns Hopkins Hospital): From the results presented it would appear that the virus particles in many cases are increasing in number. The question is: How do they increase as morphologic elements? In your situation, it would appear that a virus makes contact with the plasma-cell membrane and a new virus is grown. Is that right?

Dr. Beard: Dr. Gey wants to know what you think about the mechanism of replication, how the virus increases, and what is occurring at the membrane.

Dr. de Harven: No information is available concerning the lymphoma virus particles but for the leukemia particles it would appear that what is seen is probably an ultimate step in the duplication of the virus material.

Dr. Gey: Let us consider, then, another matter. In all the studies on viruses in terms of morphologic display, is there any evidence of simple binary fission? Is it possible even to speculate on how the virus grows as individual elements-in the cell or out of the cell or anywhere else?

Dr. de Harven: We have no evidence of binary fission at all. That is a proviral stage which has thus far evaded our morphologic techniques.

Dr. S. A. Bose (University of Illinois): In the last 2 days we have heard about some marvelous techniques for the study of the interaction of viruses in cells, viruses in tissues, and viruses in the whole animal, including immunologic methods and the examination of the tissues by microscopy. What have all these to do with cancer? I ask this provocative question for this reason: Very few cells during the process of carcinogenesis actually become malignant. How do you examine the process in some cells, when they are surrounded by other cells in which that particular and specific process is not actually occurring?

Dr. de Harven: You mean the exact distinction between a cancer cell and other cells?

Dr. Rose: In examination of the cells by any technique described, how do you know you are examining the specific process by which a cell is likely becoming malignant? You are examining the ititeraction between virus and cells, but you are not examining the particular process by which the odd cell becomes malignant. Consequently, the last 2 days have been a fine seminar on virology; I cannot say in any way that it has been a seminar on the study of carcinogenesis.

Dr. Beard: I do not think you should go quite that far.

Dr. Bose: Let me put it this way. When one feeds an animal azo dyes, one gets certain changes in the liver. One of those changes is a loss of glycogen from the liver cells. Not many people would say that the loss of glycogen, which occurs in every liver cell, has something do to with this specific process by which the one cell of many becomes malignant; this is the problem you face with your virus interaction with cells and tissues.

Dr. Beard: No, this is not the problem under consideration here. At the time we undertake to talk about tumor viruses and their relationship to the cell, we have already passed the point at which we know about the malignant state of the cell under consideration. For instance, in myeloblastosis, there is no doubt that the cell being studied is malignant. Actually, the nature of the biochemical processes that cause the cell to become malignant is an entirely different problem remaining for study.