1918] Gardner: New Pacific Coast Marine Algae Til 465 



larger colony is built up, according to the species, the cells all remain- 

 ing within a eonnnon tegument. Tlien at maturity each small cell of 

 the colony divides into a few gonidia, the whole mass of cell walls 

 dissolves thus setting free the gonidia. The other line of development 

 starts likewise with a single cell. This continues to increase in size 

 without division until maturity is reached when it is many times larger 

 than the vegetative cells in the other method of development, after 

 which the whole protoplast divides into gonidia, either by successive or 

 by simultaneous divisions. These questions need further observa- 

 tion and study before any decision of value can be made concerning 

 them. 



Pleurocapsa gloeocapsoides Setchell et Gardner sp. nov. 



Plate 39, figs. 15-17 



Strato molli, nitido, gelatinoso, sordido, 1-2 mm. crasso; division- 

 ibus cellularum regulariter in directionibus tribus ; cellulis singulis 

 globosis, in coloniis sub pressione angulatis ; cytioplasmate 4-8/^ diam., 

 homogeneo, dilute coeruleo-viridi ; parietibus propriis conspicuis, fus- 

 eescentibus; coloniis 2-8-cellularibus tegumento amplo, homogeneo, 

 hyalino, molli, gelatinoso indutis; gonidiis 2-4 in cellulis vegetativis 

 ordinariis formatis, 2.5-3. 5/i, diam. 



Colonies associated into soft, glistening, gelatinous mas.ses, 1-2 mm. 

 thick : cell divisions regularly in three planes ; cells globose when 

 single, angular from mutual pressure in colonies; protoplast 4-8/i, 

 diam., homogeneous, pale blue-green ; cell wall conspicuous, brownish ; 

 colonies of 2-8 cells enclosed in an ample, homogeneous, hyaline, soft, 

 gelatinous tegument ; gonidia 2-4 formed in unchanged vegetative 

 cells, 2.5-3.5)Lt diam. 



Growing on a water-soaked log, at the nuirgin of a salt marsli. 

 Alameda, California. Type no. 4119, Gardner. 



(fleocupsa crcpidinum Collins, Ilolden and Setchell, Phye. Bor.- 

 Amer. (Exsicc), no. 1151 (not of Thuret). 



The material of the above meiitioiud distribiitioii was collecled in 

 191.'! ill tlie same locality in which the tyi)e nuiterial of /'. (jlovocup- 

 soidcs was collected in October, 1917. The gonidia, if present in tlii" 

 original collection, were overlooked, and the close resemblance t)f 

 the species in the vegetative stage of Thuret 's Glococapsa crcpidinum, 

 so well and amply figured in Bornet and Thuret (Notes Algol. I, 

 1876. j)l. 1), led the authors of the above mentioned exsiccatae to 

 place it in that species. Now that well foi-med, typical gonidia are 

 found in great abundance it seems best to I'emove it from the Chro- 

 ococcaceae and place it in the Chamaesiphonaceae, if we are to adhere 

 to the well recognized distinction l)etween these two families. The 



