10 PROCEEDINGS OF THE ACADEMY OF 



points of resemblance between them. These are (1). " The large open median 

 nodular space." 



(2). The infra-marginal, dark, and more or less broken line or lines, with separa- 

 tion and distortion on the peculiar wavy longitudinal striation at that point." 



(3). " The sharp, clear, parallel, transverse strirn. 



This last character seems to me, to afford a valuable clue for determining the 

 primary or secondary nature of any undulate or apiculate form. As a general 

 rule, it will be found, I think, that parallel transverse strice, coinciding with 

 an undulate margin, imply a secondary or metamorphic action on a species 

 whose original outline was smooth; while on the other hand, a radiant trans- 

 verse striation, or, rather, one constantly perpendicular to the marginal line, indi- 

 cates that an undulate outline mprimary. Thus, e.^.the triundulate oatline of Na- 

 vicula Hilchcockii, concurring with & parallel, transverse, striation, is most likely, 

 secondary ; also the crenulate dorsum of Himantidium undulation (extreme 

 variety). While the large and showy Navicula Sillimanornum, (fig. 8, pi. ii.) 

 ought by the same rule to possess & primary outline, although much exagger- 

 ated and intensified. In the apiculate forms, the metamorphic force being 

 operative over but a small terminal portion of the valvular margin, renders 

 the rule less valuable. 



Although liable to many exceptions, this rule may be sufficiently general 

 to possess a practical value, when taken in connection with other means of 

 proof. 



I cannot leave these varieties of N. firma, without adverting to the species 

 known as 2V. Hilchcockii, which has been regarded by some, as a variety of 

 the former species. There is a triundulate variety of iV. firma approaching 

 very near to it in size and outline, but differing essentially in the median 

 line, which in N. Hilchcockii is remarkable as having on either side a double 

 line, nearly parallel to its course throughout. That it is of the habit of iV. 

 firma, however, can hardly be questioned; and although not clearly traceable 

 to that species, may it not, perhaps, be a compound variety, resulting from 

 the conjugation of species reputed distinct, but in reality, only so sporangially ? 



I take occasion here to quote some very interesting remarks of Professor 

 H. L. Smith, who has long been studying the habits of living dialomacex, and 

 whose observations bearing on this and other points in that connection, will, 

 I trust, shortly be made public. In a recent letter to me, speaking on this 

 very subject of extreme variation, he says : " The variety of 2V. firma 9 like iV. 

 Hilchcockii, somewhat, on one of the slides you recently sent me (Saco River,) 

 is curious ; but if you will examine the Bridgewater deposit, I think you will 

 see something about N. follis of Ehr., that makes one almost say it is only a 

 variety of N. serians. The departure is greater, even as to form alone, than 

 that of the specimen of N. firma ? like Hilchcockii is from the type of that 

 species." He adds; " when I find N. amphirhynchus conjugating and pro- 

 ducing N. firma, Stauroneis gracilis producing St. phenic enter on, and Surirella 

 splendida, S. nobilis, quite different in form and striation, I cannot but doubt 

 the propriety of making a new species out of every different shape and 

 marking." 



The views of Prof. Smith here expressed, cannot fail to awaken a lively 

 interest in all who desire to have light thrown upon this perplexing question, 

 and it is earnestly to be hoped, that the matured result of his investigations 

 will not long be delayed. 



Navicula rhomboides is, with one or two exceptions, of all northern species 

 the most common and widely distributed. It offers a broad range of form 

 and outline. Like N. firma, it has its elliptic, produced and apiculate forms. 

 I may preface what I have to say of this species, by remarking that it is, in 

 my opinion, impossible to determine on any positive characters, which shall 

 distinguish it from N. crassinervia. The small apiculate variety which I have 



[Jan. 



