

410 THE AMERICAN NATURALIST. [Vor. XXXVI. 
plant of the northern Pacific; but there are other forms passing 
under this name of quite different habit, for which a place must be 
found somewhere else. Polysiphonia bipinnata Post. and Rupr. is 
referred to Pterosiphonia, but with exclusion of Polysiphonia califor- 
. nica Harv., hitherto generally regarded as a synonym. ‘This trans- 
fer being based on original material is undoubtedly decisive, but, 
as in the case of Rhodomela floccosa, we are left with a long series 
of forms, some of which will not go into Pterosiphonia. It may 
be that they can be included in Polysiphonia californica, but more 
study is needed. Chondria baileyana Harv: and C. sedifolia Harv. are 
restored to specific rank, but in actual collecting it is not easy to 
draw the line between the former and what is called C. senuissima on 
the northeast coast, and between the latter and C. dasyphylla. 
In regard to the Baltic forms of Polysiphonia violacea (Roth) 
Grev., which Reinke considered identical with Z. harveyi Bailey 
and P. o/neyi Harv. of the American coast, denying autonymy to 
these two species, the author considers Reinke’s identification an 
error, Harvey’s types being amply distinct from the Baltic forms. 
This is a relief to American algologists, who were about ready, if 
P. violacea and P. harveyi were united, to accept one name for all 
four-tubed Polysiphonias whatever. The union under Rhodomela 
pesos (Woodw.) Ag. of such various forms as A. /ycopodioides (L.) 
, R. virgata Kjellm., and A. rochei Harv. is possible only by 
endi an extreme range in habit and mode of fruiting. It still 
seems as if A. subfusca, taken in so broad a sense, must be an 
aggregate, to be divided sooner or later. 
A work as thoroughgoing as the present, and starting from the 
foundations, must continually reach conclusions differing from those 
of previous writers, and the author states these divergences and 
contradictions with great frankness; not ill-naturedly, but sometimes 
apparently with a little impatience with errors which could have 
been avoided by a little more careful observation. Practically 
every one who has written on or referred to the Rhodomelacee 
comes in for correction sooner or later, most of all the late Pro- 
fessor Agardh, both as the most conspicuous writer and because 
. in spite of, perhaps in consequence of, his remarkable intuitional 
perception of systematic relations among the alga, he was never a 
careful and punctilious student of the development of their struc- 
ture, and it is upon the development, rather than on the mature 
structure, that the classification of the present work is based. As 
Teei with the elaborate synonymy of Bornet and Flahault’s 



