500 Howe: PHYCOLOGICAL STUDIES 
we have identified with the Mediterranean Halimeda Tuna present 
some slight peculiarities which, however, we believe are not suffi- 
ciently constant and reliable to warrant a specific separation. In 
the first place, the American plants are more rigid and rather more 
calcified than the European specimens that we have seen. When 
rowing, they are commonly suberect and cespitose, while accord- 
g g ¥: y Pp , 
ing to Oltmanns* the flat Halimedas of the Mediterranean (‘‘ Zuna, 
platydisca’”’) have a more or less horizontal or pendulous position. 
The diameter of the peripheral cells averages about 6-12 less. 
than in the European specimens examined, though no smaller than 
in certain Adriatic representatives of H. Tuna. The filaments of 
the central strand, as shown in our FIGURES 5 and 6, are often 
inclined to cohere strongly at the node just above the points of 
fusion, while those of HY. Zuna are usually easily separable under | 
treatment, as described by Mrs. Gepp + ; however, we have seen a 
specimen from the Adriatic (leg. Titius) in which the tendency of 
these filaments to cohere at the nodes is as pronounced as in the 
plants of Bermuda and Porto Rico, and Mrs. Gepp mentions (/ c.) 
a similar condition in a specimen brought from Rangiroa by Pro- 
fessor Agassiz. The sporangiophores in the American specimens 
appear to be rather more simple than those of the European 7. 
Tuna, as we have noted above. 
The American species of the Halimeda Tuna group, thus far 
recognized, may be distinguished by the use of the following key f: 
Peripheral utricles truncate or rounded-obtuse. 
Peripheral utricles in contact for } their length or less ; utricles of the subcortical 
layer subturbinate, obconical, or clavate, 35-110 # in maximum diameter. 
H, Tuna. 
Peripheral utricles in contact for 4-} their length ; utricles of the subcortical layer 
bullate, mostly ventricose-obovoid, 110-21 5 “in maximum diameter. 
H. discoidea. 
Peripheral utricles galeate-cuspidate. H. scabra. 
Se 
* Morph. and Biol. Alg. 1: 295. f. 182; 296. 1904. 
t Siboga-Expeditie. Monographe 60: 16. oor. i 
22 Excluding Halimeda gracilis Harv., which agrees essentially with the members — 
of the Halimeda Tuna alliance in the character of the nodal filaments, but differs con- 
siderably in the form of the segments, The type of 4. gracilis was from Ceylon, but 
@ specimen dredged by the Challenger Expedition at St. Thomas, West Indies, in 5 t : 
I 5 fathoms of water, appears to agree with it in most respects, though the peripheral 
utricles are larger and more coherent and the utricles of the subcortical layer are Ve'Y 
long (mostly 300 to 700 jt). 
