On the Structure of Diatoms. 21 



the high powers used, because of its greater opacity. Deductions 

 from focal changes with reference to the various markings lying in 

 different focal planes corroborate the conclusions above expressed. 



The disks examined are on Moller's Probe-Platte, and on a slide 

 prepared by Mr. Wells. 



Hyalodiscus suhtilis Bailey. — On this beautiful little shell the 

 " engine rulings " are readily seen with almost any illumination, 

 and the inevitable concomitant of intersecting lines, whether real or 

 illusory beading is displayed. When we use monochromatic light 

 the whole scene is changed. The hyaline portion of the disk is in- 

 stantly resolved, into perfectly well defined hexagons, radiating from 

 the central nucleus. The central part, because of its greater depth 

 and complexity, is only resolved into irregularly shaped spaces of a 

 more or less hexagonal form. Every one of the five beads usually 

 seen represents the centre of an hexagonal plane exactly as in 

 Pleurosigma angulatum. 



The hexagons are well defined with a power of 7000 diameters. 

 They may also be seen with lamp or daylight. 



Triceratium favus. — The two sets of markings on this fasci- 

 nating object certainly lie in different focal planes,* and probably 

 "belong to two distinct layers." The coarse hexagonal ridges are 

 found to project from the outer or convex surface, and the inner 

 plate bears the minute markings. This is proved by the fact that 

 the fine markings show decidedly the plainest on valves that are 

 mounted with the interior surface uppermost. 



Under this superior objective the finer markings, like the larger, 

 are distinctly faveolate. Their hexagonal structure is easily seen 

 even with lamp illumination. When examining comparatively thick 

 shells, possessing a complex structure, like the one in question, the 

 necessity for avoiding errors caused by too intense or by excessively 

 oblique light becomes at once apparent. The unequal refraction of 

 the light in passing through the external silicious layer produces a 

 distorted image upon and of the interior surface. In this manner 

 distorted small hexagons may be seen along the lines of the larger 

 network by a lens incapable of clearly displaying the minute hexa- 

 gonal markings above described. 



The best results are obtained on the T. favus with a moderate 

 light nearly central. 



Surirella gemma. — This beautiful form has been subjected to 

 all the different conditions of illumination in my possession. Like 

 other relatively thick shells, the appearances presented by the mark- 

 ings vary greatly with the changing conditions of observation. No 

 trouble is experienced in bringing out the longitudinal strise, nor in 

 making the little beauty seem to " wear beads." At times the beads 

 give place to rectangles, and again after careful manipulation to 

 * See Carpenter, ' The Microscope/ 4th edition, p! 282 and note. 



