18 THE HANCOCK PACIFIC EXPEDITIONS VOL. 3 



specimens in the Herbarium of the New York Botanical Garden and 

 the fact that they are both numbered 8jo indicate that these original 

 specimens were made up from the same sample, and it was to be 

 expected that the two species described should be found mixed 

 together. 



During examination of the material from the Galapagos Islands, 

 it early became evident that many individuals fall between the limits 

 of measurement originally ascribed to the two species. Though fila- 

 ments were none too abundant in either No. 123 or No. 155, I 

 measured the width of cells of one hundred filaments in the former 

 sample to ascertain to what extent the size of cells might vary. 

 Since the shapes of cells in all filaments were essentially similar — 

 all disc-shaped, and circular in end-view and narrowly rectangular 

 or ellipsoidal in optical side-view — I assumed the linear measurement 

 of width to be a fair index of size for such a comparative study. A 

 similar plan was proposed for the type material from Porto Rico, 

 but here the filaments were so much more rarely encountered than in 



Number of filaments in Number of filaments in 

 Width of cells^ our No. 12 3 fro m the Wille No . 830b and 830e 



ill microns Galapagos Islands from Porto Rico 



4.0-4.5/x 3 



4.5-S.5/X 9 2 



S.5-6.S/>t 10 6 



6.5-7.5/x 42 10 



7.5-8.5/x 17 12 



%.S-9.Sfx 7 10 



9.5-10.5^ 9 10 



10.5-1 1.5/x 3 7 



11.5-12.5/. 6 



12.5-13.5ia 5 



13.5-14.5/x 4 



14.5-15.5/. 3 



15.5-16.5/. 3 



16.5-17.5/. 2 



Total 100 80 



2 Filaments in which the width of cells varied measurably are listed in the table 

 above according to the width of the largest cells. 



