596 WILLIAM ALBERT SETCHELL 



Xo. II (pi. 38, fig. 14) shows (A) a dichotomy with definite infra-axillary dilation 

 and (B) (pi. 38, fig. 15) the only basal fragment, with its short irregular dicho- 

 tomies, evidently a mutilated plant but one arm showing an imperfect infra- 

 axillary dilation; and Nos. 6 and 10 (pi. 37, figs. 12 and 13) show rather slender 

 fragments of no distinctive characteristics. The plates of utricles (plates 44 to 

 48), also, show a range of variation, of younger and older utricles, of more 

 slender and more stout utricles, with or without hairs, gametangia, etc. The 

 series not only shows the utricular variation among the various fragments, but 

 also within the same fragment, and even in close proximity to one another. 



Codiuin iinilaterale forma Skottsbergiamivi forma nova. 

 Plate 58, figure 16. 



Thallus cylindricus, subrobustior, usque ad 13 cm. altus et 0.5 cm. cras- 

 sus, e disco parvo irregularique oriens, primo aec[uidichotomus, mox pseudo- et 

 unilateraliter- dichotomus, dichotomiis superne gradatim curtioribus; utriculis 

 megaphysis anguste lineari-clavatis apice spherico-iumidis, usque ad cuneato- 

 obovoideis, 750 — looo [x longis, superne 65 — 350 \i. latis, membrana apicali 

 lente incrassata, usque ad 16 [j. crassis stratosisque et pilis irregulari-verticillatis, 

 prope apicem positis, ornatis; gametangiis singulis (aut subverticillatis.'), anguste 

 ellipsoideis, 300 u, longis, 75 [j. latis, supra medium utriculorum positis. 



Masafiiera: No. 8, beach drift, Loberia, Feb. 16, 1917. (Type specimen, 

 Herb. Univ. Calif. No. 549100). 



A single fairly complete plant from Loberia on Masafuera has been avail- 

 able. At first it seemed possiblj^ to be closel}' related to No. 11 B from Masa- 

 tierra, because the utricles are very similar and because of some similarity in 

 habit. Xo. 11 B, however, seems to be an eroded(.?) basal portion of such a 

 fragment as No. 11 A and to be referred to C. fer7ia7ideziamnn^ especially since 

 one dichotomy, although imperfect, shows a slight, but definite, infra-axillar}' 

 dilation. No. 8 seems of entirely different habit, and approaches so closel}-, 

 in its unilateral, subdichotomous branching, to Codiiini unilaterale Setch. et 

 Gard. (Proc. Calif. Acad., 4 ser., 12: 710, pi. 15, figs. 30, 31, and pi. 36, 1924) 

 of the Gulf of California that there seems no character by which to segregate 

 it except its seeming antipodal geographic discontinuity. In habit and dimen- 

 sions, the Masafuera specimen seems true C. iinilatcralc\ in utricle dimensions 

 it still seems within the limits of the species. Its extreme antipodal discon- 

 tinuity of distribution, however, seems worthy of being recognized, at least by 

 a form name. 



Xone of the Juan Fernandez members of the subgenus Schicocoaiiivi is, 

 in our present knowledge, known from the western (or even the southeastern) 

 coasts of continental South America. Brief notes are appended for each of the 

 species occurring on the west coast of South America and not on those of 

 Juan Fernandez. 



