372 



KNO\VLi:i)Gl-:. 



OCTOBEK. 1912. 



form, Cnscinodisciis ccntnilis, where we seem to the corners of the hexagons are stuck little bosses, 

 arrive at the region of imperfect resolution : that is, slightly projecting, seen mostly at the right hand 



the finer detail. Th 



structureless covering over 



the hexagons is here arranged 



in a symmetrical pattern. 



reminding one of a tesselated 



pavement. Interspersed are 



the somewhat irregular 



|)atches of bright dots, well 



defined as to the outer ones. 



but running together am! 



indistinct in the centres. .\ii 



aperture of l-4() would 



render them plainer, but 



even then some specimens 



defy resolution altogether. 

 Another and interesting 



form just on the borderland 



of resolution b\' an oil- 

 immersion is Aiihicoiiisciis 



Kiftonii, a very difficult 



object to photograph, because 



of the inequalities of the 



surface, and a striking picture 



is imijossible. There is a 



figure of this diatom in Plate 1 of Dallinger's 



" Carpenter," taken by Mr. Nelson with a low- 

 power. A draw- 

 ing also of the 

 finer structure 



FicrKK 411. AitliicuiliscHs Kittonii. frotii 



another valve. Objective plate and screen the 



same, magnification two thousand seven hundred 



and fifty dian)eters, exposure half an hour. 



FlGURK 412. I'ihtWs oi Plcurosigma 

 forniostiiii, X 1750. 



accompanies :i 



second coininuni- 



catiou to the 



yuekett Club In 



Messrs. Nelson and Karop, in 1887. There ing over the gap, as 

 seem to be no published photographs of the finer in Figures 407 and 

 structure, but an at- 

 tempt is now made in 

 Figure 410 to supply this 

 deficiency, taken at one 

 thousand nine hundred 

 diameters. Only over a 

 small part of the picture 



side of the print and at 

 the top. .\bove all appears 

 one of the processes 

 (there are four in A. 

 Kittonii) characteristic of 

 this genus. 



b'igure 411 is from the 

 fragment of another \alve, 

 with smaller cellules but 

 rather bolder areolations, 

 taken at two thousand seven 

 hundred and fifty diameters. 

 No attempt is made here to 

 include the bosses, though a 

 few are indicated, as black 

 spots, where the object is 

 out of focus. At the top 

 of the valve (not shown 

 here) the surface is ruptured, 

 producing a crack, and one 

 of the broken edges shows 

 the same zig-zag appearance, 

 which may be said to be 

 produced by a fibril. The 



evidence is positive which we could scarcely 



have expected to find 



amongst such fine 



structure. In another 



part of the valve 



the hexagons are 



torn away, leaving 



the rings of minute 



areolations project- 



FlGLKl. 41j. 



Fibrils of P. Iialticum, X 1750. 



408 of the much coarser 

 details. One might, 

 indeed, in some of these 

 forms map out the spaces 

 and produce an atlas. 

 One might, indeed, go on 

 to infinity producing 

 examples, were it not 



are the areolations in focus, and not good even that space, and the editors, forbid. The Figures 

 at that. It should prove inter- 412-417 shew a number of fibrils taken from 

 esting, however, as presenting different diatoms of various size and contour. 



Fic.t Ki. 415. 



Single fibril of T. 



favHs, X 3000. 



Figure 416. l-'ibrils m :,. 



discus unr.iNflhn.L;, 



FlGl'Kl-: 417. Fibril of Aiiltjco- 

 discus Kittoiin. X J750. 



the general character of this form, th.nigh it And here the writer tliinks it is time to stop 

 cannot be produced upon the same plane. At saving his breath tor a future occasion— perhaps . 



