In Periplaneta americana, the large American roach, bacteriocytes are found in both sexes. These bacteriocytes are scattered throughout the fat tissue and their cytoplasm is filled with microorganisms. Evidence is presented to show that the intracellular forms are diphtheroidal bacilli. These diphtheroids are transmitted from one generation to another through the ova.
A method is described whereby tissues, free from contaminants, may be obtained from Periplaneta. A medium and a new method, the "spotting technique," are described, by means of which initial cultures of the parasites were obtained from eggs within capsules, from ova and fat tissue bacteriocytes. Two conditions appear necessary, one to initiate adaptation to the new environment, the other to bring its complete realization to fruition. When development has been properly initiated, further adaptation occurs with successive transfers until finally certain other media appear suitable. Approximately 14 per cent of the isolation and cultivation attempts succeeded. All of the isolations were studied and it was found that three morphologically distinct types had been cultivated. In general one host yielded only one type, but one female revealed three types and another two. One type was also isolated once from a fecal emulsion. The probable reason for this result is discussed.
The three types isolated were diphtheroidal bacilli resembling one another closely enough to be considered a single species but invariably offering distinct minor differences to warrant a separation into three distinct types. These three types have remained true to their original forms and sizes through 52 transfers. The sizes, general morphology, and tinctorial reactions of the three types cultivated correspond to the intracellular parasites of Periplaneta americana. The cultural and biochemical activities of the strains did not reveal any sound characters for differentiating types. Serologically, however, useful distinctions were found.
Some additional evidence along immunological lines is offered to show that the microorganism cultivated is a representative of the identical species parasitic within Periplaneta americana.
The evidence appears sound to the writer that the Periplaneta parasite has been isolated and cultivated over 20 times and that it is a bacterium belonging to the genus Corynebacterium (the diphtheroids). For the species the name Corynebacterium periplanetae nov. sp. variety americana is proposed.
The three cultural types did not produce forms small enough to pass through the pores of Berkefeld "N" candles.