Coscinodiscus astcromphalus. By T. W. Butcher. 723 



this valve. It is sufficient to indicate that in the areolations neares 

 to its centre, divisions occur within the secondaries into two or more 

 tertiary spaces, caused apparently by the deposition of minute 

 particles of a siliceous nature, and small circlets are formed. From 

 these, smaller branches are pushed out in all directions to join other 

 circlets forming at irregular intervals in other secondaries in these 

 perforated caps. In this way a fine veil is woven over the outer 

 surface of the mature valve, constituting an outer strainer to engage 

 the coarser particles suspended in the water in which the diatom 

 lives. This sediment, as it is deposited, will be easily removed 

 by the movements of the water itself. 



The function of the perforated caps is suggested as being that of 

 a fine filter, the water passing through the polygonal cells, of which 

 they are the outer protection, on to the openings in the basal 

 membrane, which, distinguished by the name of the "eye-spot 

 layer," forms the floor of the cells, and is the inner surface of the 

 valve. But Mr. Nelson suggests that a new lens, referred to later, 

 has enabled him to see that these openings are fitted with a still 

 more finely perforated membrane. The perforations here are so 

 fine that Mr. Nelson for once does not give an exact measurement, 

 though he does say that if one assumes that the length of the 

 tubule is four times its diameter, one might occupy one's time 

 counting a hundred a minute, for ten hours a day, for the space of 

 1381 years, allow 10 p.c. deduction for packing cylinders in a cube, 

 and then one would have the correct number for filling a box 1 c.mm. 

 in size. This can be conceived as an interesting and useful exercise 

 for the mathematician who might be with Mr. Nelson when he 

 says that " we are apt to talk volubly about 1000 and 2000 diameters 

 without in the least realizing what we are talking about," and, I 

 suggest, neglect the fact that it is only by assuming that these 

 tubules are passing through the silex at the rate of 60,000 per 

 inch, and that their diameters are equal to the interspaces between 

 them, that one can give the measurement of the diameter of each 

 tubule as jooWo" m -> or 0'2116 micron. As an example showing 

 what figures can be made to do, it is excellent, but as scientific 

 micrometry it is of as little value as the diagram supplied would 

 be to the artist engaged on the suggested model, unless he had 

 previously read a note in small type on another page,* which says, 

 " The next inner circle of black dots should be represented about 

 the same size as those at present in the peripheral circles, and, like 

 them, of irregular shape." A better illustration of the advantage 

 as a scientific record of a print made from a " straight," unmanipu- 

 lated negative, over the sketch of an imaginative draughtsman, as 

 skilled with his pencil as with his figures, need not be sought. 



To the biologist, however, a more pertinent question arises. 



• See this Journal, 1910, p. 698. 



