242 Evans: AIR CHAMBERS OF GRIMALDIA FRAGRANS 
It is clear from a comparison of transverse, longitudinal and 
horizontal sections that a distinction may be made between the 
dorsal layer-of air chambers in Grimaldia and the more deeply 
situated layers. In the dorsal layer the original chambers show a 
secondary partitioning by a system of more or less vertical cell 
plates, the free margins of which sometimes bear scattered teeth, 
apparently always less than two cells long. Except for these 
teeth the chambers lack filaments completely. In the more 
deeply situated layers, the chambers are much simpler and rarely 
show evidences of any kind of outgrowth. These conclusions 
show the incorrectness of certain statements made by Stephani, 
K. Miiller, and Massalongo and the essential correctness of Schiff- 
ner’s account. 
The complex conditions found in the green tissue of Grimaldia 
are duplicated by Plagiochasma and by certain species of As- 
terella. The other genera showing the Reboulia type of air cham- 
ber have a much looser green tissue, the secondary partitioning 
being-less highly developed or absent altogether. 
ORIGIN AND ENLARGEMENT OF THE AIR CHAMBERS 
The development of the air chambers in the Marchantiales 
has aroused a good deal of interest among students of the Hepati- 
cae, and the history of the subject is fully given by Barnes and 
Land (1). The explanation which they advance to account for 
the origin of the chambers differs in certain respects from the older 
explanation advanced by Leitgeb and accepted by many of his 
successors. Leitgeb’s explanation was based primarily on his 
study of Riccia, but he extended its application to the more com- 
plex genera. According to his ideas the air chambers do not 
originate in compact tissue, and no splitting of cell walls is involved 
in their formation. They arise, rather, on the surface of a young 
thallus and are due to a cessation of upward growth in certain 
jimited areas, the surrounding parts growing upward vigorously. 
The areas where growth is supposed to cease are situated in most 
cases where four of the surface cells come together; they mark the 
lower ends of the chambers, the vertical extent of which depends 
upon the degree of upward growth which the surrounding parts 
exhibit. 
