94 On Transmitted Illumination. 



■which Was taken after the tumbler had been cut in half, so that 

 the posterior side should have no influence over the illumination 

 or the appearances presented by the anterior surface.* 



Now I think every one who will impartially look at the copy 

 of this photograph of the tumbler taken under the conditions 

 of illumination which I have mentioned, must admit that it is 

 impossible from such appearances to determine the true struc- 

 ture ; and to show how small an influeuce may operate upon 

 the mind, in coming to any conclusion, it will be found that 

 merely altering the direction of the illumination exactly reverses 

 the appearance — this change resulting entirely from a very 

 slightly unequal illumination of the hemispheres which conveys 

 the idea of shadow, and consequently elevates or depresses the 

 markings according to the direction of the light by which it is 

 examined. 



I do not bring this instance forward as a proof of what the 

 structure of any one diatom valve may be, I believe it would be 

 most hazardous to draw any conclusion whatever : for my own 

 idea is, that a piece of glass covered with hemispherical depres- 

 sions, or even perforated, might be made to produce very 

 similar appearances. 



It may reasonably be asked, what value can attach to such 

 an illustration as this tumbler affords, for the purpose of deter- 

 mining the peculiar and minute structure which no doubt 

 belongs to the diatom valve ? I can only give my opinion that 

 it must be taken as negative evidence, and I would recommend 

 those observers who are anxious to determine the correct struc- 

 ture of the siliceous diatom valve, to take into consideration its 

 function, and consequently the formation that that function 



* The cut should be viewed as it stands, and then when turned upside down. 

 A piece of paper having a hole in the middle three quarters of an inch square, and 

 laid over the engraving, conceals the true form, and makes it look exceedingly like 

 a portion of a diatom valve. 



