This investigation was originally suggested by the research of Leuchtenberger, Schrader, Weir and Gentile ('53) and by subsequent related work (Leuchtenberger and Schrader, '55; Leuchtenberger, Weir, Schrader and Murmanis, '55; Leuchtenberger, Schrader, Hughes-Schrader and Gregory, '56; C. Leuchtenberger, R. Leuchtenberger, Schrader and Weir, '56; Leuchtenberger, I. Musmanis, L. Murmanis, Ito and Weir, '56; and C . Leuchtenberger, Weir, Schrader and R. Leuchtenberger, '56). In their research a deoxyribonucleic acid (DNA) analysis was made by the method of microspectrophotometry in the visible region of the spectrum, mainly on semen sperm and testis germ cells of men of known fertility and suspected infertility, and subsequently, to a limited extent, on the semen sperm of bulls. By this cytochemical technique strong presumptive evidence was obtained that the DNA content of the sperm is related to fertility and infertility. Final published conclusions were that there might be three types of DNA deviation in the sperm of infertile men, "persistently low," "persistently high" and "fluctuating" in contrast to a high degree of constancy in the sperm of men of proved fertility. Analysis of testis material confirmed the results obtained from semen sperm. It was concluded that a precise amount of DNA seemed to be necessary for the sperm to fulfill its function of fertilization. Since morphologically abnormal sperm were excluded from analysis, the DNA deviation found was separated from that factor. Also, no evidence was found that the DNA deviation had any relation to sperm count or motility. The independent type of DNA deviation seemed to prevail. On the other hand, it might be that the inclusion of too few samples of poor count and/or motility among the total analyzed might have had a similar effect on finding a correlation between DNA deviation and count and motility as the exclusio~ of morphologically abnormal sperm had on eliminating the DNA deviation associated with abnormal morphology. It is also possible that failure to carry the analysis far enough might be responsible for missing a statistical conclusion blurred by exception and experimental error in too few samples. A third possibility is that non-correlated and correlated DNA deviation might occus in different degrees in different types of material.This type of research is of obvious interest, not only because of the large human implications, but also because of the possible application to animal breeding. At the same time, because of the widespread prevalence of the theory of DNA constancy in the cells of a given species, and, consequently, the belief that it is, at least in part, the hereditary material, the evidence that DNA could vary quantitatively, and vary meaningfully, was of considerable