1 32 THE BOTANICAL MAGAZINE. LVoi. xxxiv. No. m. 



The vessels were placed facing the north window of a room in which 

 the temperature varied during the severe cold of winter from 0° to 

 20°C. in 24 hours. In testing fronds to be cultured under the micro- 

 scope, we carefully rinsed them with brushes to clear them of foreign 

 matters, especially putting aside the young frondlets which are often 

 germinating on the mother fronds. The fructified marginal portion 

 of the frond thus tested was torn up into small pieces and placed in 

 vessels containing sea water to a depth of 1—2 cm. Throughout the 

 investigation the pieces were immersed in the culture fluid without 

 being taken out of the water twice a day, so as to imitate the ebb of 

 the tide. The water was changed only at long intervals, being changed 

 but two or three times during the period of study. 



Spores were soon liberated from the marginal portions of the 

 pieces. Some of them at first took an amoeboid alternation of shape 

 (Fig. 1) and soon after became roundish. After a period of 2-3 days 

 many of the spores began to germinate, taking the appearance of 

 a short process (Fig. 2), and in a week almost all had germinated. The 

 short processes continued to elongate further and further and when 

 they had become much grown, almost 19 days after germination, had 

 developed into slender filaments which were divided here and there into 

 branches and were divided into many segments (Fig. 3, 4). Most of the 

 spores liberated from the fronds sank to the bottom and crept along by 

 elongating filaments. Those hanging in the slimy mass of the original 

 floating frond also prolongated filaments, which is evidence that settling 

 of the spore is not essential to the prolongation of filaments. 



In some cases two or three filaments were emitted from different 

 parts of a spore and often from the two opposite poles (Fig. 4c). 

 In hunger cultures the contents of some of the original spores became 

 almost empty, the wall being lined with thin layers of protoplasm 

 (while injothers this was not the case), but the chromoplastid became 

 fainter in colour as the culture grew older. In the nutrition cultures 

 the cavity of the original spore was full and the chromoplastid was 

 vividly colored. The filamentous cells were full of chromoplastids when 

 short, but when elongated the cells proximal to the spore contained 

 chromoplastids and the distal ones were faintly colored or almost 

 colorless both in hunger and nutrition cultures. (Fig. 46, 6, 8, 9). 



At the beginning of the study we took the filaments for bodies 

 like the protonema of moss ; but observating that the distal ends 

 were always colorless, in both kinds of cultures, and that the fila- 

 ments were usually longer in the hunger than in the nutrition cultures, 



