Polysiphonia of the Tropical Pacific, I — Hollenberg 
57 
monly arise endogenously from the prostrate 
branches. The chiefly prostrate habit has often 
in practice been almost, the sole criterion for 
distinguishing Lophosiphonia from Polysipho- 
nia . Kylin (1956) emphasized another feature 
characteristic of many species generally placed 
in the genus Lophosiphonia, a feature which 
was included in the original description of the 
genus. This feature, the dorsiventrality of the 
shoot apex, is well exemplified in L. cr is tat a 
Fa.lkenbe.rg, in which trichoblasts are unilateral. 
However, dorsiventrality may be determined 
by various factors in addition to the habit and 
manner of origin of erect branches from pros- 
trate branches, or the unilateral branching of 
erect branches. In certain species of Polysipho- 
nia short and more or less determinate erect 
branches arise cicatrigenously from spirally ar- 
ranged primordia resembling scar cells on the 
prostrate branches. Such prostrate branches are 
certainly not basically dorsiventral in spite of 
the seemingly dorsiventral habit. Furthermore, 
in certain species of Polysiphonia the first 
formed pericentral cell of a given segment is 
directly distal to or above the first formed peri- 
central cell of preceding segments. This is one 
type of dorsiventrality (unilaterality), and in 
such species the tetrasporangia likewise develop 
in a straight row rather than in a spiral arrange- 
ment as in most species. However, in species in 
which the tetrasporangia are in nonspiralling 
rows, the trichoblasts, if present, are spirally 
and not unilaterally arranged. In such cases one 
is at a loss to decide whether the apex of erect 
branches is radial or bilateral in construction 
unless one designates the determining feature or 
features. 
Accordingly, the writer is presently of the 
opinion that the genus Lophosiphonia should 
include only those species which have the fol- 
lowing features: (1) chiefly prostrate habit, 
(2) endogenous origin of erect, mostly simple, 
and more or less determinate branches from the 
prostrate branches; and (3) dorsiventral or bi- 
lateral apex of all branches, as evidenced by 
unilateral origin of either lateral branches or 
trichoblasts on the erect branches, or both. 
In the following account when the positions 
of trichoblasts are described as "one per seg- 
ment ” this indicates an arrangement in which 
more accurately a trichoblast or a scar-cell occurs 
on each segment not bearing a branch. It should 
also be noted that branches sometimes appear to 
arise in connection with trichoblasts when in 
fact they arise, as is frequently the case in P. 
tuherosa, in a delayed exogenous manner from 
the primordial cell which bears the trichoblast 
before the latter is shed. 
The following symbols indicate the chief col- 
lectors of the materials studied: Da., E. Y. 
Dawson; D., Maxwell S. Doty, University of 
Hawaii; G., Malvern Gilmartin Jr., University 
of Hawaii; IF, G. J. Hollenberg; JL, T., and 
B. , G. Hollenberg, Roy Tsuda, and R. Buggeln; 
L., C. R. Long, graduate student at the Uni- 
versity of Hawaii; and T., Roy T. Tsuda, grad- 
uate student at the University of Hawaii at the 
time. 
All D. collections were made by Maxwell S. 
Doty unless otherwise indicated. Further eco- 
logical data concerning collections at the Uni- 
versity of Hawaii are given in his notebooks. 
Most of the dredged materials from Hawaii 
with Doty numbers were from collections made 
by Tetsue Matsui on the Pele Expeditions spon- 
sored by Mrs. Mary Eleanor King and the Ber- 
nice P. Bishop Museum. 
Relatively few dried herbarium mounts of 
Polysiphonia from the areas concerned are avail- 
able for study. Furthermore, it is very difficult 
to make dependable determinations of most of 
these small algae from dried materials. Hence 
exsiccati have been studied to only a limited ex- 
tent and mostly in the study of types or other- 
wise authentic material. Herbarium specimens 
examined other than types are indicated as fol- 
lows: bish, Bishop Musum, Honolulu, Hawaii; 
C, University of California, Berkeley. 
The chief areas represented in the study are 
treated in the following sequence with the fol- 
lowing approximate latitudes and longitudes: 
Midway Island (28°N, 177°W); Hawaiian Is- 
lands (19-28°M, 155-176°W) ; Johnston Is- 
land (16°N, 169°W) ; Line Islands (2-7°N, 
157-l62°W); Marquesas Islands (10°S, 
140°W) ; Tuamotu Archipelago (25— 16°S, 
130-1 50°W) ; Phoenix Islands (5-7°S, 170- 
174°W) ; American Samoa (13-14°S, 168- 
170°W); Fiji Islands (15-21°S, 178°W to 
176°E) ; Gilbert Islands (2-3°N, 172- 
177°E) ; Marshall Islands (4-l4°N, 161- 
