272 Bacillariece 



inch 1 . The well-known deposit at Richmond, Virginia, U.S.A., is 

 very extensive and reaches a thickness of 30 ft., while on some of 

 the recent geological surveys beds have been discovered in the 

 western states of America no less than 300 ft. in thickness. These 

 earths contain on an average 80 / of Silica. 



It is generally assumed that the earliest appearance of fossil 

 Diatoms is in the Upper Cretaceous (chalk), although Castracane 

 has recorded the occurrence of certain species in coal from the 

 English Carboniferous beds, and Edwards 2 has stated that he has 

 found valves of Diatoms belonging to the genera Synedra and 

 Melosira in still older rocks in New Jersey. These observations, 

 however, require confirmation. 



The Bacillariese has been placed by some authors in close 

 proximity to the Conjugate and by others as an order of the 

 Phseophyceae, but the characters of Diatoms are sufficiently dis- 

 tinctive and their structure so uniform as to warrant their position 

 as a distinct class, the affinities of which are doubtful. 



I have adopted, with slight alterations, the classification of 

 Diatoms put forward by Schiitt 3 , and since followed by Lemmer- 

 mann and others. It is to my mind the most natural one, as it 

 separates all those Diatoms with a radial symmetry from those in 

 which the frustules are zygomorphic or otherwise irregular. 



Classifications based upon the disposition and mode of division 

 of the chromatophores, such as those suggested by Pfitzer 4 , Petit 5 , 

 Pelletan 6 and Ott 7 , are impracticable owing to the fact that so 

 many genera and species are unknown in the living state. That 

 published by Prof. H. L. Smith 8 and subsequently adopted by 

 Van Heurck in his ' Treatise on the Diatomaceae,' is based upon 

 certain features in the structure of the valves, such as the form of 

 the raphe, but these characters are not so clear and distinctive 



1 In a report on the chemical composition of the plankton of the Baltic Sea, 

 Brandt states that 675,000,000 of the dried frustules of Diatoms (mostly Chcetoceros) 

 weigh one gramme. Cfr Brandt, ' Beitr. zur Kenntn. der chem. Zusammens. des 

 Planktons,' Wisseusch. Meeresuntersuch., Neue Folge, Bd iii, Heft 2, 1898. 



2 Edwards in Amer. IVIonthly Micr. Journ. xx, 1899, p. 292. 



3 Schiitt in Engler & PrantPs Die Naturl. Pflanzenfam. I Teil, i, Abteilung b, 

 1896. 



4 Pfitzer, ' Untersuchungen iiber Bau und Entwicklung der Bacillariaceen,' 

 Bonn, 1871. 



5 Petit in Bull, de la Soc. Bot. de France, Paris, 1877. 



6 Pelletan in Journ. de Micrographie, xvi, 1892. 



7 Ott in S. B. k. Akad. Wissensch. Wien, cix, 1900. 



8 H. L. Smith in The Lens, Chicago, 1872. 



