The basic nature of viral selectivity for certain species, certain organs, certain cells, and certain sites within these cells constitutes a broad and deep problem that penetrates to the level of the biochemical composition and interaction of the parts of host cells as well as of viruses. The problem is not special or new to students of tumor viruses, for from the beginning it has attracted and repelled, stimulated and discouraged, and confused and bemused investigators of virus diseases.

Logically, it has been supposed that the specificity of viruses and of response to them is in some way related not only to the genetic, but also to the epigenetic, or differentiative controls of cells. One naturally looks to experimental embryology for help, but so far it has proved that material entities concerned with morphogenesis and differentiation in higher animals have been more difficult to isolate, preserve, and analyze than have viruses. Julian Huxley (1) recognized the situation as recently as 1958: "The method by which tissues and organs differentiate in the course of normal development is at the moment the main blank space in biology's map."

As it seems to be a general principle in scientific advance that a field with a higher head of pressure transfers some of its potential energy to one with a lower head (provided the water gate is open), it should follow that virology currently may be in a position to increase the depth of the morphologic "mudhole." Burnet (2) was alert to this circumstance when he wrote that "... it is less widely recognized that the behavior of viruses may be equally relevant to the understanding of the differentiation of parts and morphologic control within the cell." One of the purposes of this presentation is to propose that the converse may also be true.

It may seem strange that the polyoma-tumor complex should be selected as a subject for studies dealing with sensitivity and specificity of response. Of all presently known tumor viruses, this agent appears to be the least discriminated against by the largest variety of specialized cells, at least in the mouse. It should be noted parenthetically that, although the term polyoma has found wide usage, the word itself lacks a certain amount of specificity, for before this agent was recognized it was established by investigators of avian leukoses (8,4) that the viral agents of these diseases can induce multiple types of tumors, and might be referred to as polyvalent. It will become apparent during the course of this presentation that the polyoma virus, like other viruses, shows definite signs of limitations of action in many parameters, including those relating to species susceptibility, strain susceptibility, tissue susceptibility, and type of response by different cells as well as by cells of the same morphogenetic lineage at different levels of development. Because they offer breadth in variety of types and location of cell-virus interaction while yet retaining specific limitations, the several polyoma virus-tumor systems represent a versatile set of probes, usable for exploration of many facets of tumor-virus phenomena.