1911] SNOW—EPIPHYTIC ALGAE 365 
varied from 0.0028 to 0.006 mm. in diameter; while from a3 
to I per cent reproduction decreased, and the length and number 
of filaments diminished. This reduction in length was accom- 
panied also by a great increase in the diameter of the cells and the 
assumption of a spherical shape, followed by an almost complete 
fragmentation of the filaments previously formed (fig. 15). The 
diameter of these cells reached as much as 0.0228-0.028 mm. 
In this condition a well defined reaction for cellulose was obtained 
in the membrane with the chloriodide of zinc, but this was not 
obtained in the filamentous stage. In some cases a strand of 
protoplasm was seen to extend from each chloroplast toward the 
center; this probably extended to the nucleus, though it could 
not be determined definitely. 
The alga reproduces by means of zoospores 0.005-0.0065 mm. 
long and 0.0018-0.003 mm. broad (fig. 18). They are pointed 
at both ends, have a single chloroplast, a single cilium, and a pig- 
ment spot. They are very amoeboid in their motion, bending 
back and forth, and changing their shape from time to time. 
After a period of activity they come to rest, assume a spherical 
form, and germinate immediately. The zoospores are formed by 
the successive division of the contents of a mother cell into 2, then 
into 4, 8, 16, or many segments (fig. 17). Usually there are 16 or 32 
from each cell. They may be formed from any cell, but they seem 
to be formed first in the central cells of a complex rather than 
in the ends of the filaments. They are liberated by the gelatinifi- 
cation of the whole membrane and its expansion by the motion 
of the zoospores until one by one the zoospores become liberated 
and swim away. Often the process is very slow. 
In respect to the zoospores, as well as to the vegetative structure, 
the resemblance to Conferva and Botrydiopsis is manifested, as all 
three have amoeboid zoospores with a single cilium; a difference, 
however, lies in the fact that these zoospores have only one chloro- 
plast, while the others have two. 
The mode of germination of these zoospores and the develop- 
ment into filaments seem to be the same as occurs in the genus 
Stigeoclonium. Soon after a spore comes to rest (fig. 19), it 
divides into two, and then into a chain of four, the two outer 
