All adult urchins collected in nature contained dividing specimens of Biggaria gracilis, but only 3% contained those of Entodiscus borealis and Madsenia indomita. Urchins were starved for 2-3 weeks; dividing ciliates were absent. They were then fed generously on Laminaria for 5 days, and the condition of the ciliates was recorded at intervals. In most of the urchins, division began in E borealis after 10-15 hours and in M indomita after 50-60 hours. In all the urchins, it resumed in B gracilis after 20-30 hours. Thus, division occurs continuously in B gracilis, except in drastically starved urchins. Although it occurs discontinuously in E borealis and M indomita, it is doubtful that cycles of division are inherent in them. It seems more likely that their division requires the copious ingestion of suitable bacteria, whose numbers become adequate only when abundant food is present in the urchin gut.