FOSLIE—LITHOTHAMNTA. 191 
however, essentially, different from that species, peculiarly in structure. It is actually 
most nearly connected with Lithoph. craspedium. ‘This is particularly the case with 
f. obpyramidata, which is sometimes hardly distinguishable from a Lithoph.. eras- 
pedium f. abbreviata feebly developed, the latter sometimes showing as thin branches as 
those of f. obpyramidata. In structure, however, there is some difference between the 
two species, as the medullar hypothallic cells of Lithoph. craspedium form a row of 
short cells, rather regularly alternating with one or two or up to three rows of long ones, 
whereas in the species in question the corresponding long cells, as above mentioned, 
frequently form more rows. When these rows of elongated cells occur in numbers 
of six to seven, a longitudinal section sometimes gives the impression of a faint cup- 
shaped stratification, as is usual particularly in branching species of the genus Litho- 
thamnion. The form subhemispherica is vigorously developed and much larger than any 
of the other forms. The branches are a little thinner, more crowded, and more curved 
than is usual in the typical form. A single specimen only of this form is known 
from the reef of Egmont Atoll. There are also forms of Lithoph. Gardineri which 
bear in habit rather close resemblance to Lithoph. craspedium f. subtilis. The latter, 
however, is a form as yet insufficiently known, which may perhaps represent a 
separate species, but, being in structure in all essentials consistent with Lithoph. 
craspedium, is for the present to be regarded as a feebly developed or stunted form 
of this species. 
Lithoph. Gardineri seems to occur abundantly on reefs, and particularly associated 
with Lithoph. craspedium and Goniolithon frutescens, sometimes coalescing especially 
with the species last mentioned. Specimens are also found coalesced with corals. 
No doubt it is an important reef-builder, though perhaps less so than Lithoph. craspediwm 
and Lithoph. onkodes. 'The species seems to occur most numerously and best developed 
in the Chagos Archipelago, particularly on the reefs. Small specimens have been 
brought home from “ Nullipore shoal in lagoon” of Egmont Atoll. However, it is also 
found well developed at Coetivy. But some of the specimens brought home from this 
locality have no doubt grown under conditions different from those at Salomon Atoll, 
partly appearing as a well-defined f. obpyramidata. 
The species is scantily furnished with ripe sporangia in the month of September. 
Most of the specimens collected on Salomon Atoll in May—June are sterile or provided 
with empty conceptacles. 
Localities. Chagos Archipelago: Salomon Atoll, reef, f. ¢ypica, f. obpyramidata ; 
Egmont Atoll, reef and lagoon shoal, f. typica, f. subsemispherica, f. obpyramidata ; 
Coetivy, E. reef and other localities, f. typica, f. obpyramidata (Stanley Gardiner). 
fh 104 | 
