814 SUMMARY OF CURRENT RESEARCHES RELATING TO 



outside tlie grasp of the ordinary achroraatics. The result of this 

 increased power in my hands was to enable me to split up the supposed 

 one plate of silex forming the valve of P. formosum into three, and thus 

 add two more vertical notches to the standard by which we measure 

 our objectives. 



The advantage of applying such increased power to the elucidation 

 of minute structure generally is so evident, that it is only necessary for 

 me to j)lace the existence of the compound structure and its character 

 beyond a doubt to leave the matter in your hands to apply for yourselves. 



When I had the honour of bringing this question before you twelve 

 months ago, I was met by the objection that the appearances I described 

 were diffraction effects — meaning false effects— and was asked if I had 

 examined the diatoms mounted in a dense medium as well as when 

 mounted dry. After the exhaustive manner in which diffraction has 

 been discussed within the last twelve months, and the modification of 

 opinion to which that discussion points, I do not think it necessary to 

 meet the first objection ; but on the second point I may say that I have 

 since examined a slide of Pleurosigma formosum mounted in phosphorus, 

 and found all my previous opinions confirmed. There has also cropped 

 up from time to time the objection that the interference of light coming 

 through a grating, and the impossibility of separating two such gratings 

 — if they existed — from each other must vitiate any conclusions that 

 might be drawn from mere visual appearances. I recognize the force 

 of the last objection, but at the same time beg to point out that within 

 certain limits it applies rather to the old dry objectives of narrow aper- 

 ture than to the new oil-immersions. With the latter the depth of 

 penetration is so little that if two layers are separated by ever so narrow 

 an interval, for all chance of interference they might as well be a mile 

 apart. Of course, in asserting this I am supposing a large central cone 

 of light, as being the only correct method of illumination with such a 

 glass, the slightest deviation from which will produce error. But even with 

 an oil-immersion of wide aperture it is still possible for two layers to be so 

 closely connected that interference occurs, and no doubt, under such cir- 

 cumstances, it would be impossible to be sure of the structure. Had no other 

 method been adopted by me than to record an appearance as true simply 

 because it appeared such under the Microscope, I should deserve all the 

 censure you could apply to such a method of working. Such, however, has 

 not been my method, but when there has been the slightest doubt, I 

 have formed no definite opinion of any structure until seeing it isolated 

 from everything which could interfere with the definition. Thus three 

 layers of structure have been figured by me in Pleurosigma formosum, 

 because I have been enabled to isolate them, but I have never ventured 

 upon describing more than two in the other species of this genus, 

 although one might be led by analogy to suppose there were three. 

 Leaving out the question, then, of the middle layer in. the finer forms as 

 one on which I can offer no direct evidence, the task is much simplified 

 when trying to prove the existence of two layers in Pleurosigma an- 

 gulatum, it being necessary only to deal with the two opposite sides of 

 the valve. On looking over a spread slide of this diatom, mounted dry, 

 we at once discover that different valves present different optical ap- 

 pearances, and on further examination shall also find that the different 

 valves have different curves, and that the same curves and appearance 

 always belong to each other. The prints here to-night marked 1 and 2 



