1903.J HEDOPIIYLLUM SPIRALE, SP. NOV. 1G9 



form which was satisfactorily reported by the previous authors. The 

 rhizomes around the secondary stem grow more and more, issuing dendritic 

 ramules, each of which ending in a small ha|)ter. The primary stem grows 

 larger adding considerable length and tliiclmess ; and bending suddenl}^ 

 downward at the base. The scar of the primary lamina faces against 

 the substratum as a rule. (fig. 11). 



When the secondary lamina grew to its fall size, the margins at the 

 transition region begin to expand so as the characteristic auricle now ap- 

 pears. At the same time bulla tion takes place in the lamina at the upper 

 part of each auricle. The scrolls are formed at the margins of the auricles, 

 the tertiary laminae are sprung out from them and the secondary lamina 

 gradually wears away leaving its scar as before. 



At this stage the secondary stems begin to grow larger, adding their 

 thickness and gently expanding upward. When the secondary laminae 

 have wholy disappeared the secondary stems become wedge-shaped having 

 their narrow ends connected with the primary stem, and provided with 

 the tertiary frond at each corner of the broadened ends. 



The tertiary rhizomes appear at the bases of the tertiary stems. The 

 wedge-shaped secondary stems bend downward so that the tertiar}^ rhizomes 

 are thus greatly helped to easily access the substratum, just as the primary 

 stem has done before. 



The successive growth takes place once in a year repeating the similar 

 process at each growth. The branching mode is, as readily be understood 

 from the above given description, a regular dichotomy. But after several 

 years the old stems wear away and some of the branches are detached ofP 

 in the heavy storms during winter. The final result is an irregularly 

 dichotomous, highly complicated stolon, with stripe-form lamina at each 

 ultimate point. This form has been fully described^^ and familiar to us. • 



The above relates the development of Arthrothamnus bifidus. In 

 ArthrotJiamnus Kurilensis the entire plant is erect, not as De Toni has 

 doubted. ^^ Consequently the dorsi-ventrality of the stems are not clearly 

 manifested. The primary stem and holdfast are persistent and the succes- 

 sive holdfasts or the rhizomes are not normally found. In other respects 

 as the former species. 



Had the new species been legitimately reckoned under the genus 



1.* Among the descriptions of Arth, bifidus, the one by Mertens, Linnea. 1829. p. 55. 

 under the name of Fucus cornucopiae. shall be recommended. 

 2. De Toni : Sylloge Algarum. III. p. 369. 



