This discussion has included only a partial list of the systems now under development at Oak Ridge as part of the feasibility studies for the Molecular Anatomy Program. It is evident that we are still in the "Robert Goddard" phase of this work. It may not be premature, however, to suggest several conclusions. Biomedical scientists are discouraged on discovering that developmental efforts cost more, by one or two orders of magnitude, than pure research. In part this is because the full cost of development is generally shown, while in pure research some of the costs may be hidden, or the funds supplied by several sources. Regardless of the reason, the fact remains that development is expensive, as is well understood in nuclear physics and space science. The role and mission of the large national laboratories, and the kinds of research that should be done in them, have been discussed by Weinberg (63). The studies described here were in part stimulated by his ideas. We have been unable to find an environment outside a large national laboratory where a program like the Molecular Anatomy Program could be undertaken at the present rate. It appears that programs which attempt to make use of the multidisciplinary approach characteristic of national laboratories should be carefully designed and should evolve experimentally. There is less chance of success when a program is an administrative invention than when it evolves from scientific invention and discovery. It has been pointed out (64) that most program decisions in science are secret decisions in the sense that the scientific community as a whole does not participate in them. If a choice is to be made at some future time between large-scale expenditures for exploring space, for developing new weapons systems, for constructing new accelerators, for designing large reactors, or for systematically developing methods to explore the molecular basis of human disease, then we will need sufficient information to evaluate each alternative fully, and the information should be generally available. It appears desirable, therefore, to allow the Molecular Anatomy Program to proceed to a point where the full range of its contributions and its inherent limitations may be seen. A rational choice may then be made.