Pseudopod activity or hypervillosity of the cell surface is consistent in tumor cells, whatever their origin. These phenomena represent a cell reaction of general significance and cannot be considered as characteristic when observed in the absence of viral particles. The same may be said of the hyperplasia of the endoplasmic reticulum, which can be marked in some virus-infected cells, especially in those grown in vitro (fig. 22). Incredible amounts of small vesicles can often be seen in relation to the membrane surface, and their dilatation into vacuoles can be followed (fig. 23). Whether this is a sign of "micropinocytosis" and corresponds to ingestion of the nutrient fluid medium or whether it bears a relation to viral particles cannot be settled from static pictures.

What then is the value of ultrastructural morphology in cancer research? It will be seen in the second part of this survey that far from being deceptive, electron microscopy has become a fundamental tool. If there have been described cell reactions to which one can attach little or no specific meaning, the discussion, nevertheless, has been by no means without a significant bearing on the problem for the following reasons:

1. This paper was meant to constitute a basis for discussion rather than a review.

2. It is important to describe in detail the structures indicated only to avoid conferring on them interpretations of particular significance.

3. We are able to stress the absolute similarity, as far as ultrastructure is concerned, between virus-induced tumors and human cancers.

4. The one surely distinctive positive sign of cell involvement remaining to the ultrastructural morphologist in the study of tumors is the presence of the virus itself.

Although this last assertion must seem quite "obvious," this is not actually true, as will be appreciated from further consideration.