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



these lines has no specific or varietal significance whatever. 

 Almost every gathering shows valves both with and without 

 them, and innumerable specimens exhibit an intermediate con- 

 dition, i.e. where the lines are more or less broken, or where they 

 are present on some of the primary areas of a valve and not on 

 others. They are scarcely ever complete, but generally stop 

 short of the umbilicus, as in the var. balearica. Valves without 

 them are otherwise identical with those possessing them, having 

 exactly the same range of variation in other respects, and this 

 applies equally to the so-called A. JIdlleri. 



While reliance on such characters as the foregoing leads to the 

 improper separation of allied forms on the one hand, it tends in 

 other cases to the opposite error. Thus several varieties of 

 A. glabratus have been described, and while some are, as before- 

 mentioned, only smooth valves of A. splendeiis, there are others 

 which, so far as I know, cannot be identified with any special 

 form of that species, and which may probably be themselves 

 entitled to specific rank. A. vulgaris also, as generally under- 

 stood, includes forms which have really no close relationship. 

 One such form is nothing but A. undulatus^ as it is found in 

 Redondo Beach and other deposits, with mostly fourteen areas. 

 The deposit mentioned contains numerous valves of the ordinary 

 form, with six areas, a few with eight, ten and twelve, a good 

 many with fourteen and a few with sixteen and eighteen. The 

 structure of these is absolutely identical with that of the six- 

 rayed forms, and it is as absurd to separate them as it would be 

 to separate forms of A. Heliojjelta with six areas from those with 

 more. Other forms commonly ranked under A. vulgaris are 

 simply valves of A. adriaticus with the pseudo-raphes wanting, 

 as already described, while others seem to be similar, but with 

 deeper and more abrupt undulations. The undulations in 

 A. adriaticus are very shallow, so much so that Grunow origin- 

 ally described it as flat ; but in view of the considerable variation 

 in this respect found in the valves of A. undulatus and other 

 species, the character would seem to be of doubtful importance. 



Probably the nearest approach to a really flat condition is 

 found in the three-sided A. mari/landicus, in which the six areas 

 show a very slight difference of level near the centre only, else- 

 where blending with each other imperceptibly. This species has 

 a more or less distinctly three-sided umbilicus, and appears to be 

 identical with the Symbolophora trinitatis of Ehrenberg. Ralfs 

 has argued against this view on the ground that >S'. trinitatis is 



