NOTES ON DINOFLAGELLATES 17 



organisms multiply rapidly, no division stages were found, even 

 when the organisms were kept under frequent observation 

 from 5.30 a.m. to 8 p.m. Difficulties in obtaining satisfactory 

 light for microscopic work limited this observation to the 

 hours of daylight. 



In April, 1921. however, observations were made from 

 8 p.m. onwards, the species used being A. herdmani. It was 

 found that division began soon after dark, but was arrested, 

 at whatever stage it had reached, as soon as the organisms 

 were exposed to light. Attempts were made, by the use of 

 " Ratten " colour filters, to observe, without arresting, 

 division ; but, although red, yellow, blue, orange, green and 

 purple lights were tried, the effect was always the same. The 

 organisms, at whatever stage of division they might be when 

 the light was turned on, continued swimming actively for an 

 hour or so, after which they became more and more quiescent 

 and, after about another hour, broke up. Thus, by keeping a 

 reserve of material in the dark and taking new samples from it 

 every half -hour or so, most stages of the division were observed 

 (Figs. A to D), but the whole process of. division was not 

 watched in any one individual. 



Division in A. herdmani is oblique, and thus typical of the 

 Dinoflagellata, but owing to the small size of the epicone, it 

 appears to be almost longitudinal (Figs. A to D). The schizonts 

 remain together swimming actively for some time after 

 division, this stage being presumably homologous with the 

 " chain," which is best developed in the Peridinioidae. By 

 3 a.m., most of the schizonts had separated, and no division 

 stages were found after that hour. 



Division has not yet been observed in the other forms 

 described below. 



The curiously constant distribution of the various forms 

 suggested transplantation experiments, but, so far, these have 

 yielded only negative results. 



