Knowledge. 



With uhich is incorporated Hardwicke's Science Gossip, and the Ilhistrated Scientific News. 



A Monthly Record of Science. 



Conducted by Wilfred Mark Webb, F.L.S., and E. S. Grew, M.A. 

 SEPTKNn^KK, 1911. 



THE TRUE STRUCTURE OF THE DLATOM 



VALVE. 



By T. F. SMITH. 



(Coiitiiiuai from page 293.) 



Figure 14 is from the inner side of another valve, 

 the first ever seen and taken, showing the fracture 

 through undoubted perforations. The strongest 

 argument hitherto for " beads " in the liner forms 



aKva\'S ran between them, 



was that the fracture 

 as the weakest part. 

 Mr. E. M. Nelson, 

 at the discussion in 

 the Ouekett Club, 

 gave this testimony : 

 "The difficu]t\' has 

 al\\a\'S been felt that 

 in foniiosiiiii the\" 

 had never been abh' 

 to see the holes : but 

 no\\' it was show n 

 that there was. ini- 

 doubtedly. a per^ 

 forated membrane : 

 the\' could take it 

 and see for them- 

 selves that it was a 

 perfectly plain thing; 

 indeed, if this was 

 not real, then he 

 could onh' sa}" that 

 all other things were 

 hocus pocus. With 

 their very best lenses 

 thev had tried to find 



out about fonnosiim. Mr. Smitli had found this 

 fracture, had shown it to him, and that at any rate 

 the fracture did run through the holes." This 



Figure 14. The inner side of P/c';(;-osi^/)!rt fonnosmn 



showing the fractures through the holes, the first ever 



seen under these conditions, X 1750. 



here not in any spirit of self- 

 glorification" on the part of the present writer; but 

 rather that so valuable a testimony should be 

 added to strengthen his own. 



Particular attention is also directed to this print, 



as it exhibits a 

 «te^ structure of squares 



instead of the cus- 

 tomar\" round beads. 

 This is in accordance 

 with the theory of 

 Dr. .\bbe. expressed 

 in his paper on the 

 relationship of aper- 

 ture to power. He 

 shows there that all 

 the optical details of 

 an object developed 

 w ith insufficient aper- 

 ture to resohe them 

 properlv, whether 

 square, 1 oze n ge- 

 shape, or triangular, 

 are depicted either 

 oval or round. 

 Figure 4, from the 

 same side, where the 

 valve has not been 

 acted upon, may 

 seem to contradict 

 this view . The present writer, however, has already 

 given his opinion that these appearances do not 

 denote structure, but are only a collection of focal 



331 



