Suspensions rapidly frozen to –196°C. were in all cases innocuous, whereas those frozen slowly were capable of transmitting leukemia. The infectivity of slowly frozen material varied from an estimated 0.0001 per cent to 1 per cent of that of fresh material, and this figure probably represents the percentage of surviving leukemic cells.

Particles of spleen and lymph node reacted to slow and rapid freezing in the same manner as suspensions prepared from them.

For one of the strains rapid thawing was less injurious than slow thawing; for the other two the rate of thawing seemed to be immaterial.

Infectivity was equally well preserved after freezing to –21°C. whether freezing occurred spontaneously after supercooling or was initiated near the freezing point by inoculation with ice, or whether thawing was slow or rapid.

Suspensions already slowly frozen at temperatures of –2° or lower, whether spontaneously or by inoculation with ice, could no longer be completely inactivated by subsequent rapid cooling to –196°C. Unfrozen suspensions initially above the freezing point or supercooled to –2°C. or –8°C. and then rapidly cooled to –196°C. were inactivated. This protective action of previous slow freezing was most marked when the initial temperature of the frozen suspension was –15°C. or lower; when it was –2°C. protection was barely detected.

These observations indicate that the changes which are peculiar to rapid freezing alone and lead to complete inactivation take place during rapid transition from the liquid to the solid state, in a range of temperature lying between –15°C. and the freezing point. Temperature measurements carried out in this range showed that suspensions were about equally infections whether the temperature at their centers dropped from 0°C. to –15°C. in 30 minutes or in 1 minute; when the drop occurred in 12 seconds or less, the suspensions became innocuous.