INTRODUCTION. XI 



19. Rhabdonema arcuatum, Kiitz., November 1849 and March 1852. 



Plate E. fig. 305. 1.-III. 



20. Melosira nummuloides, Kiitz., October 1852. Plate XLIX. fig. 



329./". 



21. „ Borrerii, Grev., March 1852. Plate L. fig. 330. 



22. „ subflexilis, Kiitz., March 1852. Plate LI. fig. 331./. sp. 



23. „ varians, Ag., October 1851. Plate LI. fig. 332./ sp. 



24. Orthosira orichalcea, W. Sm., October 1855. Plate E. fig. 337. 



I.-IV. (from Thwaites). 



25. Encyonema prostratum, Ralfs, April 1852. Plate E. fig. 345. 1. & II. 



26. Colletonema subcoh^rens, Thw., October 1847. Plate E. fig. 353. 



27. Schizonema Grevillii, Ag., July 1849. Plate E. fig. 364. 



If we add to the above list Navicula Jirma, Kiitz., and Suri- 

 rella bifrons, Kiitz., as recorded by Dr. Griffith, together with 

 Cocconeis Pediculus, Kiitz., Cymbella Pediculus, Kiitz., and Am- 

 phora ovalis, Kiitz., detected by Mr. Carter, we shall have in all 

 thirty-two species belonging to seventeen genera, which have 

 been observed in a state of conjugation. These numbers are 

 small in proportion to the whole of our ascertained British 

 forms, and leave much to be accomplished by future observers. 

 One reason for the paucity of observations on this process in 

 the Diatomaceae is no doubt to be found in the changes which 

 usually take place in the condition of these organisms at this 

 period of their existence. During conjugation the progress of 

 self-division is arrested, the general mucus-envelope or stratum, 

 produced during self-division, is dissolved, and the conjugating 

 pairs of frustules become detached from the original mass ; 

 they are thus more readily borne away and dispersed by the 

 surrounding currents, or the movements of worms and insects, 

 and their detection becomes in consequence more casual and 

 difficult. By far the greater number of the species I have men- 

 tioned belong to those genera whose frustules are adherent or 

 attached by stipes to foreign bodies, or which form continuous 

 filaments or aggregated frondose expansions. Not more than 

 four, viz. Cyclotella Kiitzingiana, Navicula jirma, Amphora ovalis, 

 and Cymbella Pediculus, are to be regarded as free forms : the 

 reason I have just given will account for this circumstance, and 

 the larger proportion of adherent or frondose species detected in 

 conjugation may doubtless be ascribed to the firmer position 



