July, I'JIC] 



yENDO.—NOTIJS ON ALQJIi NEW TO JAPAN, V. 



269 



^-. 



haptcrs do not appear in them. A.s a rule, the rhizome is limited 

 to the bavsal part of frond but a lower l)ranch on an erect shoot 

 may frequently modify to assume a similar appearance (Fig. 2). 

 Ch. umhellata H. ET H, is described from such form. 



A. and E. S. Gepp combined Ch. gelklioides Hary. and L. 

 catenata Hakv.. into one species though with some doubt. 

 After comparing the type specimens in the Herbarium of the 



Trinity College, Dublin, I have 

 but to agree with their view. 

 The}' remark about the modes 

 of ramification in both species. 

 But in the well-grown typical 

 forms, the lateral branches are 

 always opposite. It may varj'' 

 as alluded to above and never 

 be mentioned as a specific dif- 

 ference. Okamura pointed out 

 that the specimen from Sj'dney 

 which he hesitatino^lv identified 

 with Ch. gelidioides Harv. has 

 loosely "intricated branches by 

 coaliseing to each other." In 

 Japanese forms of what has been passing as L. catenata, similar 

 fusion of branches is frequently met with. The point of attach- 

 ment, as far as I have observed, is mere thickening of cell-wall, 

 the lamellar structure of which being more or less distorted at 

 the point. No special hapters or tenancula has been found there. 

 In the Agardhian Herbarium Ch. gelidioides Harv. is repre- 

 sented by a slender specimen with sparing patent branches. It 

 resembles in general aspect to Gelidium crinale, hence, very likely, 

 the specific name. Similar form is also to be found on our coast, 

 especially among the northern inhabitants. There are many 

 intermediate forms to link it with the large typical form. 



In short, the plant has a very wide distribution in the 

 Pacific Ocean and is highly variable in its external appearance. 

 Locality. On the coasts of middle and southern Japan. 

 Distribution. Australia. 



