A series of crosses between a naturally-occurring self-compatible accession ofLycopersicon peruvianum and a closely-related self-incompatible accession were used to demonstrate that the mutation to self-compatibility is located at the S-locus. Progeny of the crosses contain abundant style proteins of about 30 kDa that segregate with the S6and S7-alleles from the SI parent and the Sc-allele from the SC parent. The S6and S7-associated proteins have ribonuclease activity whereas the Sc-associated protein is not an active ribonuclease. This finding indicates that S-RNases are determinants of self-incompatibility in the style and that the ribonuclease activity is essential for their function.

The name B. coli mutabile was employed by Neisser (1906) to designate a non-lactose-fermenting coliform organism which suddenly acquired the capacity to ferment lactose. The mutation-like variation became manifest through the formation of secondary colonies on lactose agar. Subcultures from a secondary colony proved to be active fermenters while similar cultures from the remainder of the colony had no such capacity. Other workers have since shown that this mode of variation occurs in a wide variety of bacterial species and that various carbon compounds may be involved. Since the term "mutabile" denotes variation by mutation, Burri (1910) objected to it and proposed the name, "imperfec-0 122 80 74 12 4

The first detailed description of a method of endospore formation seems to be that of Robert Koch (1876). Since that time numerous workers have studied this problem with the result that various conflicting methods have been described. The present state of general knowledge concerning the cytology of endospore formation is revealed by comparing the accounts given in recent reviews and textbook literature. While there is lack of uniformity in the descriptions two conflicting methods are quite generally described. These may be illustrated by brief quotations.Conn (1926) says:Within the rod which is to bear the spore a small round body appears which is more highly refractive than the rest of the rod and does not take stains as readily. It grows until it becomes as broad or even broader than the parent organism and perhaps half as long. Then the remains of the original organism die axvay. This in brief concise form is the view expressed by Robert Koch (1876) for B. anthracis and repeated by DeBary (1887) for B. megatherium. It is still to be found in many modern textbooks of bacteriology and Cook (1932) in a recent review states that the method described by DeBary could not be bettered.A different method is clearly stated by Buchanan (1930) who wrote as follows:The endospore formation in a bacterial cell is first indicated by the appearance within the cell of highly refractile granules. These tend to collect at one point in the cell and unite to form a larger globule. This increases in size and density and finally surrounds itself with a firm mem-133 on July 31, 2020 by guest http://jb.asm.org/ Downloaded from I. M. LEWIS brane or spore wall. The spore seems to be made by the contraction of the protoplasm with extrusion of water.

The complicated life cycle described by Lohnis and Smith (1916, 1923) for Azotobacter has not been fully confirmed by subsequent investigation. Jones (1920) who had previously reported reproduction by gonidia, accepted the doctrine of symplasm but was not able to confirm conjunction, the formation of endospores, or filtrability. Mulvania (1919) confirmed reproduction by budding and gonidia. de Regel (1932) accepted the gonidial method of reproduction and filtrability but rejected conjunction, symplasm, and transmutation of cells. Beauverie (1925) failed to confirm any of the salient features of a cyclogenic life history. In view of the conflicting results obtained by various workers

No abstract