340 



PROCEEDINGS OF THE ACADEMY OP 



Considered as generic, these characters may be classed as essential and non- 

 essential. The former of these divisions embraces such structural characters 

 as median and terminal nodules, central lines, alse, costse or canaliculi. These appear 

 to bear to their parent organism a relation somewhat similar to that held by the 

 tracheae and stigmata of insects or the nutritive vessels of plants in their 

 own sphere. Among them the susceptibility to variation is notably less than 

 in the latter or non-essential class, in which may be placed such shifting and 

 superficial characters as size, external configuration, striation, &c. These ought 

 more properly to be regarded as mere accidental phenomena, constantly mod- 

 ified by agencies depending on climate, locality, soil, and the mineral consti- 

 tution of the water which contains the diatomaceous growth. 



This tendency to extreme variation in these non-essential characters is for- 

 cibly illustrated everywhere throughout this country by the strange modifica- 

 tions of size, shape, striation, punctulation, and even condition of aggrega- 

 tion, observable in such species as Navicula firma and 2V. rhomboides, under 

 differing conditions of locality and reproductive agencies.* 



Assuming, then, that all these variations in size, shape, striation, &c, when 

 observed in Diatomacea?, are non-essential phenomena, let us review the facts 

 relating to the species under consideration. We have here a form, which, by 

 virtue of the possession of two of the most invariable and essential of generic 

 characters, viz., alse and canaliculi is allied to Surirella, one of the most 

 ancient and widely distributed of known genera, and to Nitzschia, Synedra 

 and Amphipleura, more modern and weaker genera, by the non-essential char- 

 acter of external configuration, and in the case of Amphipleura by the occa- 

 sional presence of submarginal ridges, a resemblance only observable, how- 

 ever, in the smaller or sporangial brood ol S. intermedia and S. anceps, n. sp., 

 hereafter to be described. To Nitzschia, the likeness ot these smaller frus- 

 tules is more marked, and I am disposed to think that the form figured in 

 Prof. Gregory's paper on Fossil Diatomacese. (Trans. Mic. Soc. No. viii. 

 Mic. Journal,) as Nitzschia sigmatella, may be an undeveloped speci- 

 men of the present species. From Nitzschia, however, the absence of any 

 trace of keel, together with the fact of the arrangement of the marginal 

 puncta upon the same plane, sufficiently removes S. intermedia. From these 

 facts it seems probable that this curious diatom is in reality a transitionary or 

 comprehensive type species which, along with other forms in the gathering, 

 themselves comprehensive, departing more or less widely from the typical 

 Surirella and tending towards the Synedroid type, has resulted under certain 

 peculiar and exceptional circumstances while the modern peat was being 

 deposited. 



To explain the apparent anomaly of the presence and perpetuation of this 

 assemblage of influences in this particular locality, will be my endeavor 

 when proceeding with the description of the species, although I am fully 

 aware that my premises, founded on the presumed geological relations of the 

 Saco deposit to the sub-peat and to the surface soil are, as previously stated, 

 not fairly proven. 



As a preliminary, I may state those conditions which appear to be essential 

 to the developement and fixity of this and other intermediate forms constitu- 

 ting so unique and eccentric a grouping in the Saco mud.f These may be 

 assumed to be : 



* My friend Prof. H. L. Smith, of Ohio, informs me that he has found JV. rhomboides in the con- 

 dition of a Colletonema. 



f It may be urged that in the foregoing remarks on these species and their relations to the 

 typical Surirella, I have 1 advanced the idea of gradual generic transmutation, rather than the 

 true thiory of the comprehensive type, which implies no such serial progression as is apparent in 

 the present instance. To this objection I would observe that, while no such serial progression is 

 implied, neither is it inconsistent with this theory ; for the intermediate type differs from a sub- 

 genus, nut only because it does not transcend the essential characters of genus, but for the reason 

 also, which makes it superior to ordinary species and varieties, viz., that it possesses pecu- 



[Dec. 



