82 REV. E. O^MEARA. 



contents divide and afterwards re-unite, alternately in the 

 case of A. longipes, directly in that of A. subscssilis. Ac- 

 cording to Liiders, there is always found a gelatinous sheath 

 surrounding the infant cells, which force themselves out by 

 an opening at the end. 



It appears confusing that two species so nearly related 

 should exhibit such different conditions in the formation of 

 their spores, and therefore the author expresses a wish that 

 observers residing near the sea-shore will carefully examine 

 fresh sj)ecimens with a view to ascertain satisfactorily the 

 process of spore-formation. 



After the Achnanthidece Pfitzer ranges the group Coc- 

 coneidea, in which are embraced the two genera, Achnan- 

 thidium and Cocconeis. Achnanthidium has been distinguished 

 from Achnanthes by the fact that while the latter is stipitate 

 the former is free. To this Pfitzer adds another mark of 

 distinction founded on the character of the endochrome plates. 

 Achnanthes has two endochrome-plates, while in Achnanthi- 

 dium lanceolatimi there is but one, which lies upon the convex 

 valve. This peculiarity places Achnanthidium in intimate 

 relationship with the next genus, Cocconeis. Cocconeis 

 Pediculus at least possesses a single endochrome-plate, occu- 

 pying a position similar to i\\&.t oi Achnanthidium lanceolatum, 

 split up on the edge, and with its scallops reaching the girdle- 

 band. It exhibits also a strong slit on one side, a circum- 

 stance which in the author's opinion shows that the Cocco- 

 neidecB are not decidedly symmetrical in the longitudinal 

 plane. This feature discovers itself in Achnanthidium like- 

 wise, in the structure of the concave valve, by a stronger 

 development of the central nodule on one side than on the 

 other. The endochrome-plate is more deeply scalloped in 

 proportion as the valve is large. The central incision some- 

 times extends so far as to effect a comj)lete division of the 

 plate. A nucleus is clearly seen in Coccoiieis Pediculus, 

 as is also a central accumulation of plasm. So the Coc- 

 coneidece are essentially distinct from the Naviculece, and ex- 

 hibit a decided analogy to the similarly epiphytic Ami^horece, 

 inasmuch as in neither does the occurrence of the longi- 

 tudinal line exhibit anything to correspond with it in the 

 structure of the primordial cell. In the Amphore<2 and 

 Cocconeidea the endochrome-plate stands related to the surface 

 of attachment. The former attaches itself by one girdle-band, 

 and upon this plane lies the middle of the plate ; the latter 

 are fixed to foreign bodies by one valve, on which the middle 

 of the plate lies. As respects the formation of auxospores, 

 in Achnanthidium it has never been discovered, but in 



