W. M. BALE ON SOME OF THE DISCOID DIATOMS. 33 



variation, the granules often forming irregular zigzag lines 

 crossing the fascicles; generally, however, the tendency is for 

 the granules of adjacent series to alternate with each other, and 

 also to be somewhat crowded, so as to form a quincuncial 

 arrangement, which in any case prevails towards the border. 

 In A. rex the alternate arrangement is much more pronounced, 

 and as the granules are crowded equally all round the markings 

 form a very regular areolation over the greater part of the 

 valve. 



The appearance of the granules themselves varies remarkably 

 in the same species. In A. Ehrenbergii some valves show them 

 in the best focus as minute, dark, sharply defined circles, while in 

 others they are more pearly, and show, much more readily, a 

 central black spot. When crowded, especially towards the 

 border, they form a distinct areolation. In A. rex the latter 

 type predominates, but near the centre the granules are more 

 pearly. In A. Barklyi and A. ellipticus they vary much as 

 in A. Ehrenbergii. And in all these species they appear some- 

 times as dark, well-defined puncta. Peragallo has figured a form 

 which he calls A. nebulosus, and which is practically a hyaline 

 valve of A. Ehrenbergii with fine puncta instead of granules, also 

 a corresponding form with the puncta arranged like the granules 

 of a typical A. Ralfsii. He thinks these valves are probably the 

 result of cleavage, of the correctness of which opinion I think 

 there can be no doubt. Corresponding forms of A. Barklyi are 

 found in hundreds in slides of that species, often so delicate and 

 colourless that they become invisible on a slight alteration of the 

 focus. How many layers has a valve of A. Barklyi% When 

 manipulating one under the microscope I saw it divide into 

 three, one extremely thin and hyaline, and another somewhat 

 thicker, but still less robust than the main disc. Here the 

 question of colour comes in for consideration, for it is probable 

 that the colour as well as the appearance of the granules depends 

 more or less on the " state " of the valve whether it consists of 

 more than one plate for instance, or whether the two plates 

 include a film of air between them. A. rex is the most brightly 

 coloured form I have seen, having the colours in sharply defined 

 zones. A. Ehrenbergii is usually blue, green, purple, or brown, 

 often showing more than one colour, but not in sharp zones. A . 

 Barklyi varies much in the same way, but is exceptionally liable 

 Journ. Q. M. 0., Series II. No. 72. 3 



