NO. 2 drouet: myxophyceae 19 



our own material that in the end I was forced to be satisfied with 

 observing only eighty individual filaments. The table above shows 

 the distribution of these measurements of cell-width in the material 

 from both the Galapagos Islands and Porto Rico. 



It is readily to be seen from the above table that there can be no 

 distinction between Cyanothrix primaria and C. Willei as described 

 by Professor Gardner on the basis of cell-width; and in neither the 

 type specimens nor in our own material can I find other morphologi- 

 cal differences. I therefore follow Professor Fremy in considering 

 that we are dealing here with a single species in which cell-size 

 varies within wide limits. That the range in size is not the same in 

 different habitats, as shown in the table above, is to be expected, 

 since the size of the individual so often depends upon the nature 

 of the environment. Professor Fremy has found the range in width 

 of cells in the collections from Algeria and Buen Ayre to extend 

 from 5.4/A to 16.3/x. In Professor Taylor's material of Nodularia ? 

 fusca from the Dry Tortugas, the range is from 6. Six to 10.3/a. 



Variation in size of cells within the same filaments also occurs 

 in all the collections seen, as shown in Fig. 4, drawn from the type 

 material of Cya^iothrix primaria, where the cells range from 9/i, wide 

 at one end of the filament to 13ju. at the other end. Such gradation 

 in size of cells from one end of the filament to the other is especially 

 noticeable in those filaments which contain the largest cells. Fila- 

 ments with large cells in the middle and small cells at the ends 

 have also been observed. Obtusely conical apical cells as shown in 

 Fig. 5 and 4. are found in all of the material seen and are well de- 

 picted in Professor Fremy's figures. 



Professor Fremy suggested (ibid.) that Nodularia ? fusca W. R. 

 Taylor should be placed in the genus Johannesbaptistla. Professor 

 Taylor obligingly sent me a part of his original material from Long 

 Key, Dry Tortugas, for comparison with our material from the 

 Galapagos Islands and with the type of C. primaria from Porto 

 Rico. The collection agrees in every respect with the type and the 

 Galapagos specimens, even to the brown-colored protoplasm in many 

 filaments and the rare occurrence, as noted in Professor Taylor's 

 diagnosis (loc. cit.J, of evident sheaths about pairs of adjacent cells 

 (see Fig. 5.). In this material the filaments are more abundant and 

 as a rule of greater length than in the Galapagos or Porto Rico 

 collections. 



