462 PROCEEDINGS OF THE ACADEMY OF [May, 



differs in the closer costse (10 to 12 in 10 jj.) and in the frequent unlike- 

 ness of the valves by the unilateral interruption of the costse on one 

 of them, though there is here some agreement in the nearly parallel 

 costse. Neither in Cleve's descriptions, nor in the figures cited by 

 him in Schmidt's Atlas 3 , does the present form find convincing presen- 

 tation. The nearest approach is in Schmidt's Atlas, pi. 42, figs. 11 and 

 12, but even in these, which are in outline an approach, there are 

 recognizable differences, beyond that very notable one, which applies 

 here as well as elsewhere — the normal grouping of the species herein 

 described into motile chains. If such grouping has been heretofore 

 observed in related species, it does not find mention in the usual 

 authoritative works, nor is a similar condition of N. viridis known to 

 me. Mounts of the diatom having been distributed to some extent, 

 it was interesting to find Mr. W. A. Terry able to send material from 

 Fall Mountain, Conn., which contained it in considerable abundance. 

 This was received in July, 1909. It showed the usual grouping, and 

 rather larger frustules than those collected by myself. Mr. F. J. Keeley, 

 who had made fine gatherings in May, at the same time and place as 

 myself, collected it also in characteristic condition during the summer 

 at Dogtown Common. Cape Ann, Mass. He also detected it in the 

 usual groups in some of his preparations of material from near Phil- 

 adelphia. Near Media it is frequent in boggy places during the 

 summer, and until the pools dry out, though seldom in much abundance. 

 It goes without saying that a history of the life-cycle of this diatom, 

 which will necessarily include an account of the formation of these 

 groups, is much to be desired. At present it is not easy even to 

 understand the process of ordinary reduplication in the case of this 

 form. If the groups do not multiply by simultaneous subdivision of 

 their constituent frustules, there would soon be irregular groupings. 

 If the subdivision is simultaneous, good gatherings should show 

 superposed groups. Neither of these conditions have been seen. 4 

 Meantime, the practical advantage of the grouping is evident enough, 

 s.:> far as concerns a motility dependent to a great degree on contact 

 of the raphe with the substratum. An isolated Navicula with con- 

 siderable breadth of valve may easily find itself disadvantageous^ 

 on its girdle, and no irregularity of the solid bottom within reach. 

 A struggle to right itself is the usual consequence of such a condition. 



3 A. Schmidt, Atlas der Diatomaceen-Kunde. 



♦Occasional groups of two or three frustules will be seen in rich gatherings, 

 and a few containing six to eight, all in one plane, have been noted. But the 

 normal number is unquestionably four. 



