Specific Diagnosis in the Diatomaceee. By Thomas Comber. 429 
Actinocyclus Elirenbergii , Eunotia Elirenbergii , and E. robusta, to 
include many forms, varying in this respect, previously regarded as 
distinct species; and his action has been generally approved by 
subsequent authors. 
Forms differing in size alone, or with only a slight difference in 
the fineness of the striation, such as Navicula rhomboides , N. crassi- 
nervia, and Frustulia saxonica, are now generally regarded as not 
constituting distinct species. They are, however, still considered by 
some to form distinct varieties, as the names of the numerous varieties 
denominated by Prof. Grunow as “ maxima,” “ major,” “ media,” 
“parva,” “ minor,” “minima,” “minuta,” and “ minutissima,” suffi- 
ciently indicate. It is probable, however, that these are, for the most 
part, mere stages of growth, especially as the formation of mega- 
frustules, by a process more or less resembling conjugation, has now 
been observed in so many species, that the increase of size thus 
attained may be regarded as a normal phase in the life-history of the 
organisms. It would, consequently, be as reasonable to describe an 
oak sapling and a full grown oak tree as distinct varieties, worthy of 
distinguishing names. 
This now generally admitted insufficiency of two of the principal 
characters formerly relied on as a sufficient distinction of species, 
renders it desirable to inquire how far dependence can be placed upon 
other characters, even yet usually accepted as of specific importance. 
The first to be noticed may be the very slight differences in the 
outline of the valve, such as are portrayed in plate vii. of Prof. Van 
Heurck’s ‘ Synopsis des Diatomees Belgiques,’ illustrating § Radiosae 
of Navicula . When we consider that forms having valves triangular, 
square, pentagonal, and hexagonal, which would formerly have been 
referred to several different genera, are by most observers now included 
in a single species of Triceratium or Stictodiscus , it is plainly question- 
able whether the very slight differences of shape, which some of Prof. 
Van Heurck’s figures show, can be reasonably held to constitute 
separate species. Prof. Gregory was the first to appreciate the varia- 
bility of outline in diatoms. As far back as 1855 he wrote : — “ The 
more that the Diatomaceae are studied the more do we perceive that 
in many species the shape or outline is subject to endless variations.” 
He therefore proposed two comprehensive species, Pinnularia varians 
and P. mutabilis , each to include several reputed species of previous 
authors. The step was in advance of the day, and Gregory’s species 
have not been generally adopted. But his view was correct, and the 
yet more extended knowledge of the present day proves that, although 
species formerly considered distinct are met with in many gatherings, 
each retaining its own character, and without any intermediate forms, 
in other gatherings they occur with every gradation of form, con- 
stituting a perfect and unbroken transition from one to the other. 
Under these circumstances it seems to me undesirable to retain 
between them the distinction of “ species.” Instances of an undue 
