When the number of a lot containing a few diseased larvae was limited to 20, 000, the following facts were revealed: The infectivity of pébrine of a lot of newly hatched larvae containing 0.5% composed of the lightly diseased and the heavily diseased was exceedingly inferior to that of the larvae containg 0.05% of the lightly diseased. Expressing the infectivity by multiplication power, it became 73.5 times in the former, while 1, 117.6 times in the latter in the moth stages. The infectivity of heavily diseased larvae is very low. The multiplication power of the disease by 6 infected larvae mixed in a lot at 3rd day of 2nd instar was only 5.7 times in the moth stage, being 0.25%. The results above are common phenomena of silkworm reared in both seasons, spring and summer. These results sugget that the difference of increment of the number of diseased moths of a lot may be derived from the following two factors: 1. Hereditarily, lightly diseased larvae can live long and repeatedly attack healthy larvae, however, when their number are few, the chance of attacking the same larvae may be few, thus increasing the number of diseased larvae during the periods of 4th and 5th instars which are the source producing diseased moths (1966), but when their number are numerous, the reversed phenomenon may be probable by the repeat infection against the same larvae, thus decreasing the diseased moths. 2. Heavily diseased larvae: Early death and very few repeat infection. It is also suggested that the optimum number of lightly diseased larvae which produce the maximum percentage of infected moths may exist at the beginning of hatching.
