Evans: AIR CHAMBERS OF GRIMALDIA FRAGRANS 241 
» left 
diit 
are shown, but the cells enclosed present a very different app 
They are not only much more numerous but are, with a few excep- 
tions, more or less firmly united, and the entire chamber is thus 
divided up into smaller chambers, some of which seem entirely 
cut off while others show their connections with other chambers. 
The seven cells shown on the right of the left-hand partition in 
FIG. 3 are represented in Fic. 4 by seven united cells, showing at 
once that these seven cells are not the cross sections of filaments 
but simply the cross sections of teeth, like those shown in Fic. 2. 
Similar conclusions would be created by comparing other apparently 
free cells in Fic. 3 with their representatives in Fic. 4. It thus 
becomes established that there are no free filaments in the cham- 
bers. It will be noted further that Fic. 4 presents a much more 
complicated condition than Fic. 3 and that the boundaries of the 
air chamber would be hardly 
distinguishable except through 
comparison with the simpler -fig- 
ure. It is probable that a sec- 
tion like the one shown in FIG. 4 
was responsible for Schiffner’s 
statements, which it certainly 
strongly supports. 
In Fic. 5 a section from 
another thallus is shown, cut at 
a still lower level. This section 
shows a loose spongy tissue, two 
of the chambers being connected 
by a passageway. Cellular out- 
growths are very infrequent, but 
a single cell, apparently free, is ae ‘ 
shown in one of the chambers, ee a ae ling 
and a single short outgrowth in pers, » 270. 
another. When compared with 
Fic. 4 the spaces are relatively larger and fewer and the tissue in 
consequence much less compact. Sections cut farther down show 
elongated spaces, similar to those represented in Fic. 2, while 
sections beneath these show the ventral parenchymatous tissue 
without spaces of any sort. 
Bt ok 
3 
