THE ANNALS 



AND 



MAGAZINE OF NATURAL HISTORY, 



[SECOND SERIES.] 

 No. 3. MARCH 1848. 



XVI. — Further Observations on the Diatomacese ; with descriptions 

 of new genera and species. By G. H. K. Thwaites^ Lecturer 

 on Botany and Vegetable Physiology at the Bristol Medical 

 School. 



[With two Plates.] 



Agreeably with the promise made in my last communication, I 

 proceed to offer a few observations upon the facts there brought 

 forward with reference to conjugation in the DiatomacetE, and 

 especially as to the bearing which these facts have upon the 

 subject of impregnation in the higher tribes of plants. 



It may be desirable perhaps, by way of preliminary, to give a 

 short general account of the phsenomena, as far as they have been 

 observed, which present themselves in the course of development 

 of a species of the Diatomacece. 



The frustules of a Diatomaceous plant, which are usually, as 

 is well known, of very definite and often very beautiful figure, 

 are continually undergoing fissiparous division — that is, the con- 

 tained endochrome of each one of these frustules divides into two 

 portions, each of which developes around itself a cell-wall pos- 

 sessing a form and character precisely similar to those of the 

 original one. The process of fissiparous division continuing, ne- 

 cessarily in course of time causes a very considerable increase in 

 the number of frustules. There appears however to be a limit 

 to this mode of propagation of the frustules, except by the inter- 

 vention of another phsenomenon — namely conjugation, or a mix- 

 ture of endochromes ; after which process fissiparous division 

 proceeds as before. 



It seems probable that physiologically we ought to consider the 

 numerous frustules which have originated from the primordial 

 frustule (the sporangium or product of conjugation) not as so 

 many individuals of a species, but rather as parts of one indi- 

 vidual, which instead of, as takes place in the higher plants, 

 cohering to form one structure, assuming forms more or less 

 modified, and exhibiting a greater or less specialization of func- 



Ann.S^Mag.KHiiit,^eT.2. Vol. i. 11 



