Evans: AIR CHAMBERS OF GRIMALDIA FRAGRANS 239 
rise to outgrowths, as Schiffner suggests may be the case. The 
original boundaries of the dorsal air chambers are not absolutely 
unrecognizable, but they are by no means as distinct as Massa- 
_ longo’s figure represents them. In Fic. 1 the boundaries of the 
chamber with the air-pore are shown at i and k, while another 
boundary is situated at I. 
The longitudinal section drawn (Fic. 2) brings out the fact 
that many of the air chambers are more or less elongated. This 
is strikingly true of those most deeply situated but is also well 
. Fic. 2. Longitudinal section through epidermis and green tissue, X 270. 
shown by the chamber with the air-pore, although the actual 
boundaries of this chamber are not definitely indicated. It will 
be noted that the upper margin of the cell plate represented, which 
extends almost longitudinally beneath the pore, is distinctly 
dentate, some of the teeth being over a cell in length. This 
accords, on the whole, with Schiffner’s statement that the marginal 
cells of the plates may project as teeth. Although some of the 
teeth shown are more than projecting cells, it would be a stretch 
of the term to describe them as filaments. The figure, therefore, 
presents no evidence of the occurrence of true filaments. Other 
sections, however, show apparent filaments, similar to those 
represented in Fic. 1. 
According to Schiffner, a section through the green tissue paral- 
lel with the surface of the thallus will at once show that the cham- 
bers are destitute of free filaments. Fic. 3 shows a part of sucha 
