W. M. BALE ON SOME OF THE DISCOID DIATOMS. 31 



exists between the two types. This agrees with the views 

 expressed by Grunow, Lagerstedt, and others ; it would seem, 

 therefore, that Peragallo is justified in treating A. Ehrenbergii 

 as at most a variety of A. Ralfsii. 



Most species of A.ctinocyclus have the markings arranged on 

 the same general plan as A. Ehrenbergii. The surface of the 

 valve is divided into cuneate areas by a number of moniliform 

 series of granules (the interfasciculate rays), which radiate from 

 the centre, or near it, to the marginal zone. Each cuneate area 

 contains a fascicle of similar moniliform series, but only the 

 central one is strictly radial, and all the others are parallel 

 with it ; and as they all stop short of the interfasciculate rays 

 they are necessarily shorter as they approach these rays. The 

 great difference in the aspect of the valves dependent on the 

 small or large number of fascicles has already been mentioned. 

 In the largest valves, where they are most numerous, they are 

 so narrow that they consist of very few series of granules, and 

 the angles which they form with the interfasciculate rays are so- 

 small that at first sight it might appear that all the series are 

 truly radial. Such is the structure in the largest valves of 

 A. Ralfsii, A. Ehrenbergii, A. Barklyi, etc., but the markings are 

 just as truly fasciculate as in the smallest form's, though the 

 fasciculi are not so patent. No amount of variation of the 

 kind described, therefore, is in itself of importance in classification. 

 But great irregularities in the arrangement of the markings 

 prevail, and there is' perhaps no other genus in which valves 

 of one and the same species present such different aspects. 

 While one valve may have the interfasciculate rays very distinct, 

 all starting from a circular central ring of granules, and all the 

 series well defined, the next may present at first sight a very 

 different aspect, owing to the denseness of the granulation, and 

 in yet another much of the appearance of regularity may be 

 lost owing to its sparseness. This is especially noticeable in the 

 centre of the valve, where there may be a regular area, with 

 perhaps a few granules in the centre, while in other cases there 

 may be no definite area at all. Usually the interfasciculate rays 

 stop short at a little distance from the centre, but in the small 

 valves of A. Ehrenbergii from Oran, as Mr. Rattray points out, 

 they cross each other. Another point of variation is the width 

 of the blank areas along the sides of the interfasciculate rays. 



