Limits of Species in the Diatomacese. By Thomas Cornier. 465 
the many living or fossil forms, as yet unknown to us, can be com- 
pared and connected with one another, that studies can be guided 
by evolution, i.e. genealogical descent, and that it will be possible 
to soundly appreciate the relative value of the forms of diatoms.” 
In Dr. Cleve’s ‘ Synopsis of the Naviculoid Diatoms ’ (1894 and 
1896), the author’s opinion as to species may be gathered from his 
remarks under Diploneis crabro : * “ This species comprises a con- 
siderable number of forms, differing in size, number of costae, breadth 
of lunulae, and in the amount or absence of constriction of the middle. 
... If only a few forms be examined, it is easy to found on them 
apparently well defined species ; but the greater the number of inter- 
mediate forms observed, the greater becomes the difficulty of finding 
any definite distinctions between them. There are all intermediate 
transitions from purely elliptical to strongly constricted forms, from 
forms with no lunulae to others with broad lunulae I have 
distinguished the following forms, which diatomists fond of species- 
making may consider as specifically distinct.” He then describes a 
number of varieties, arranged in three series, and comprising eighteen 
species of other authors. He constitutes several similar aggregate 
species, such as Caloneis Liber, C. Silicula, Trachyneis aspera, 
Achnanthes brevipes, Cocconeis dirupta , C. Scutellum , Navicula 
Hennedyi, and N. Lyra , each including a good number of species 
previously considered distinct. But Dr. Cl eve is not always con- 
sistent. In not a few cases he keeps, as separate species, forms 
described as graduating into each other. Under Navicula rliyncho- 
cepliala, for instance, he observes that this species and its variety 
Amphiceros “pass into each other; the latter graduates into N. 
avenacea and N. viridula,” yet the three species are kept up. Again, 
under Pleurosigma formosum, “ all the forms from P. speciosum are 
very nearly connected, and might be united into one single species. 
. . . Between P. formosum and P. decorum there is absolutely no 
specific difference, and by numerous varieties P. formosum graduates 
into P. pulchrum and P. speciosum ; ” yet four species are maintained. 
Such divergent views as to what should constitute a species have 
necessarily produced very great confusion, have burdened the litera- 
ture of diatoms with a vast load of synonyms, and have led, in this 
country at least, to an indisposition to study the order. When a 
valued correspondent of mine proposed to take up its study, a 
scientific friend contemptuously remarked to him, “Diatoms! you 
might as well collect and attempt to classify wall-papers.” Can no 
uniform system be adopted, which will prevent such a reproach ? It 
is quite true that diatoms are, so to speak, kaleidoscopic. Their 
characters are notoriously transitional, and have not so nearly con- 
densed round certain definite types as is the case with most other 
orders of plants. Specific limitation and specific diagnosis are corre- 
spondingly difficult. But does not this render the order more worthy 
* Part i. p. 100. 
