1922] Gardner: The Genus Fucus on the Pai-ific Coast 9 



certain emendations, using as a basis for my decision Ruprecht 's state- 

 ments {loc. cit.) supported by Farlow's contention {loc. cit.) that 

 furcatus belongs primarily to the Pacific coast of North America. 

 Borgesen (1902, p. 465) recognizes the priority of Agardh's furcatus 

 (1820) over De la Pylaie's eclentafus (1829), but rejects the name 

 furcatus on the ground, as he says, that "the specimen of this species 

 which C. Agardh described was — judging especially from his figure — 

 a small, poorly developed one (apparently a transitional form to f. 

 distichus) .'' '' The type specimen of F. furcatus is in the Herbarium of 

 J. G. Agardh. no. 00279 (Type!) Unalaska, Ex. Herb. C. Ag., accord- 

 ing to Professor Setchell. I have reproduced Agardh's figure on 

 plate 1, figure 1. and it certainly is not a characteristic portion of any 

 species of Fucus, but represents, as I think, a proliferous portion of a 

 plant. I have seen many specimens of proliferations similar to the 

 one represented by Agardh (cf. plate 2), taken from dried herbarium 

 specimens collected by Setchell at Fort Ross, the probable type 

 locality. It seems to me that the evidence is sufficiently ample to 

 warrant the assumption that Agardh 's delineation, poor and uncertain 

 as it is, refers to a plant on the California coast. 



Basis of Classification 



I am using as a fundamental distinguishing character of F. furcatus 

 as a species major the presence, usually, of a large number of cavities 

 in the fronds which are completely closed. These are, I suspect, what 

 J. Agardh (1868, p. 38) refers to under " Cryptostomata immersa saepe 

 numerosa, plurima diu clausa. " I have investigated the origin of 

 these structures and find that they originate as do the conceptacles 

 and ordinary cryptostomata, as shown by Bower (1880). A surface 

 cell near the growing point in the terminal depression breaks down 

 and the cavity is formed as described by Bower, but unlike the con- 

 ceptacles and the cryptostomata no ostiole remains. The cavity is 

 completely arched over, and no paraphyses develop. This process 

 appears to be an evolution in the direction of reduction. To these 

 cavities I am applying the term caecost&mata, as suggested by Professor 

 Setchell. to distinguish them from the true cryptostomata with ostiole 

 and paraphyses. In F. furcatus f. typicus, as I am limiting that 

 species, they are \evj abundant, up to four hundred and fifty in each 

 square centimeter, and there are no cryptostomata. In certain other 

 forms of furcatus there are a few cryptostomata perfectly formed very 

 early. The caecostomatal character seems quite effectually to mark off 



