W. M. BALE ON SOME OF THE DISCOID DIATOMS. 39 



occurs it is always the primary area, with its process, which is 

 duplicated ; we never see valves with all the areas alike and 

 having the distinctive markings of the secondary ones. 



Notwithstanding that it has been recognised that in A. undu- 

 latus the variation in question has no specific importance, being 

 found, in fact, in frustules otherwise normal, a parallel variation 

 in other cases has been made a ground for the foundation of new 

 species, even by observers as recent as Grunow and Schmidt. 

 Such instances are A. Janischii Grun., which, as I shall 

 demonstrate, is only a state of A. splendens, and A. Molleri 

 Grun., which is a form of A. adriaticus Grun. Van Heurck says 

 of A. Janischii that it " se distingue de toutes les autres especes 

 du genre en ce que la valve a toute juste moitie autant 

 d'ondulations que de divisions, de facon qu'une elevation n'est 

 suivie d'une autre elevation que pres da deuxieme appendice 

 suivant. Une espece analogue mais plus petite est V A. Jfolleri 

 d'Adelaide, qui se distingue en outre par sa structure plus 

 delicate et l'absence d'une ligne mediane." This is simply 

 equivalent to saying that each area, instead of each alternate 

 area, bears a process, and it is surprising that the writer did not 

 observe that the character referred to as so exceptional was no 

 other than he has figured in the same plate in the forma 

 sexappendicidata of A. undulatus. 



A. glabratus Grunow and A. Janischii Grunow are, in part at 

 least, forms of A. splendens, but there is a difference in the 

 relationship which they bear to that species, A. glabratus 

 simply consisting of valves wanting the secondary markings, 

 while A. Janischii is an internal disc. A. splendens commonly 

 has a distinct secondary layer showing more or less branching 

 venation, with the typical distinction between primary and 

 secondary areas, but a gathering usually includes a propor- 

 tion of valves in which the secondary layer is wanting ; and 

 although there is every possible gradation, the smooth valves 

 have been described as a doubtful species, under the name of 

 A . glabratus. Also accompanying them are valves in which all 

 the compartments bear processes, and to these the name 

 A. Janischii has been given, Janisch having figured one of 

 them (as Halionyx vicenarius) in his paper on diatoms from 

 guano. Tn Peru guano A. splendens is one of the commonest 

 species, and the typical valves, with their glabratus-iovms and 

 Janischii-iovms, are readily obtained. In a Cuxhaven gathering 

 I also find all three forms together. And in a slide of Thum's, 



