On the Classification of Diatomacese. By Paul Petit. 11 
of W. Smith are also very just. Dr. Pfitzer has likewise studied 
the plasma of the Diatomaceas very profoundly, and has published 
the result of his researches in that remarkable work, ‘ Bau und 
Entwicklung der Diatomaceen ’ (Bacillariaceen).* 
“ As the result of his researches, he introduced a new system, but 
it cannot be considered a natural one. In fact, all the groups are 
not connected naturally with each other, though the species com- 
posing the groups have a certain affinity. 
“ I must nevertheless admit Dr. Pfitzer ’s system approaches 
the nearest to a natural one, and that combined with M. Grunow’s 
it has served as the basis of the method I now propose. 
“ It is in verifying the observations of Dr. Pfitzer, and in com- 
pleting them by the study of marine and fresh-water species, which 
lie had not at his disposal, that I have been enabled to recognize 
the relation of the groups to each other. 
“ Some details are, however, necessary for the better under- 
standing the bases on which my method is founded. 
“ If we examine a large number of living diatoms, and note 
exactly the disposition of the endochrome as well as the form of the 
frustule, we shall not fail to detect the two distinct states of the 
endochrome as indicated by the Bev. W. Smith, and the constancy 
of its disposition in all individuals of the same species. We shall 
see at the same time that the connection of the endochrome with the 
frustule is common to all the species of the same genus, and some- 
times to many genera which have an analogous development of the 
siliceous envelope. We can therefore lay down the two following 
laws : 
“ 1. The internal disposition of the endochrome is alike in all 
individuals of the same species. 
“ 2. The mode of connection of the endochrome with the frustule 
is common to all species of the same genus, and frequently to many 
genera, having an analogous structural development of their 
siliceous envelope. 
“ These principles will be of great service in the classification of 
fossil species, when identification by means of their plasma is 
impossible. It seems certain to me that the connection between the 
siliceous envelope and the endochrome does not exist at the time of 
reproduction, but as observations have only been made on about 
sixty species belong to twenty-five genera, t it is impossible to 
notice this character, which, however, appears to be constant in the 
few instances with which we are acquainted. 
*• I follow Dr. Pfitzer in dividing the family of the Diatomaceae 
into two sub-families. 
* ‘ Botanische Abhandl.,’ von Dr. I. Hanstein, Heft ii. Bonn, 1871. 
t Pfitzer, toe. cit , p. 163. 
