98 TERMINOLOGY AND CLASSIFICATION OF DIATOMS. 



diameters and then reduced by phototypy to 600 diameters in the case of 

 genera (except where otherwise stated) and to 400 diameters in the case 

 of species. 



The figures have as a rule been phototyped from the original plates 

 of my Synopsis dcs Diatom'ees de Belgtque Atlas, for which they were 

 drawn from nature by Mr. A. Grunow and myself. 



The large number of new figures which appear in this work have 

 been drawn from photographs which I have made from some of my 

 preparations. As I have been unable, in the case of a few genera, to 

 procure actual specimens, I have in such cases reproduced the original 

 figures published by their authors. 



A few of the species have been reproduced from the original figures 

 of Gregory, Ad. Schmidt, etc., reduced in every case to a magnification 

 of about 400 diameters. 



The originals of all the figures in the work will be found in the place 

 cited after the name of the species. If my Synopsis Atlas is not men- 

 tioned, the figure is reproduced from the author quoted, and I have also 

 indicated by an asterisk the work from which the figure is taken. 



6. Of Genus and Species. 



In the text of the Synopsis which appeared in 1885, I gave tables 

 for the determination of genera from the work of Professor H. L. 

 Smith ; but in the interval the science has advanced, and very many 

 new genera have been established in the important works published 

 during those ten years. In another respect my ideas have become slightly 

 modified. While I certainly do not admit genera founded on unim" 

 portant characteristics, or on the variable modifications of the coleoderm, 

 such as the sheath, the stipes, etc., I nevertheless believe that reductions 

 must not be pushed to an extreme. A larger number of genera so 

 long as they are easily distinguishable is advantageous, inasmuch as 

 they restrict the number of species from which a selection has to be 

 made. 



All the forms are evolved from a small number of original forms, 

 sometimes even from a single one ; in proportion as researches multiply, 

 a greater number of links are discovered, which connect forms with one 

 another, between which there appeared formerly to be no connection what- 

 soever. 



These new links naturally diminish the value of the differential 

 characteristics previously admitted, and increase the difficulty of determin- 

 ing the species. It is therefore necessary to discover fresh divisions, 



