A count-chamber like methodes enables us to estimate the number of cysts corresponding to each developmental stage of the spermatogenesis.The multiplication of the germ cells depends upon the physiological conditions of the host: it is stopped in isolated larval abdomina, and is low in larvae of a crowded culture where pupation is delayed. In implants of 48 h old testis in. virgin adult females the multiplication rate is lower than in males and in mated females; it is as high as in situ in castrated females or in pupal hosts.The time required for the oldest cysts with spermatocytes to enter the meiotic divisions is with one exception for all these experiments the same as in situ; only in pupal hosts it will be accelerated. The subsequent spermatid differentiation in sperm also takes the same time as in situ. Thereby the premeiotic growth and postmeiotic differentiation is as efficient in all tested milieus as in situ, even when the hosts have not attained or already bypassed the metamorphosis. This autonomy is also valid for germ cells under insufficient multiplying conditions, provided that the spermatocytes have reached a certain stage in the premeiotic growth.When the first cysts reach the postmeiotic stages the further multiplication of cysts is stopped. This regulative mechanism is attributed to a specific inhibition operated by the spermatids.The autonomy of the spermatocyte differentiation and the unified behaviour of all the cells within a cyst are discussed.