63 JOURNAL OF THE [April, 



claim in Dr. Cox the first worker in photomicrography under 

 those new, and the only true, conditions. 



But, while making no claims now to be the first to originate this 

 idea of a compound structure in the finer forms of diatoms, I 

 think I may claim to have been the first to attempt to photograph 

 the different layers; not from possessing any superior powers of 

 observation, but because the work was only possible after the ad- 

 vent of the new apochromatic lenses from Germany. 



When we consider the difference between the actinic and the 

 visual foci in the best lenses of the ordinary achromatic correc- 

 tion when used photographically, and also the infinitesimal dis- 

 tances which must separate the different layers in the valves of 

 the minute forms of diatoms, it will be seen at once how utterly 

 impossible it would be to make any allowance fine enough to 

 bring the particular layer aimed at sharp in focus on the neg- 

 ative. The reduction, however, of the secondary spectrum in the 

 apochromatics entirely obliterates this difference, and what is 

 seen sharp on the focussing screen may be depended upon to be 

 sharp in the negative, given correct exposure and development. 



This capacity of standing more light was pointed out from the 

 first by Mr. E. M. Nelson, but has not received the attention it 

 deserves; and the neglect of this point has stultified the efforts of 

 many microscopists, both here and on the Continent, to get more 

 out of the new glasses than the old objectives. Unfortunately,. 

 the most flagrant examples of " how not to do it" come from the 

 very workshop which produces the glasses, and Dr. Roderick 

 Zeiss' celebrated print of P. angulatum shows how an oil-immer- 

 sion apo: of 1.3 N. A. may be made to perform no better than a 

 good dry ^-inch. 



The conclusion arrived at in my first paper is that the valve of 

 P. fonnosum — as being the one most easily studied from its coarse 

 structure — -consists of three layers, figured as Nos. i, 2, and 5 in 

 the plate belonging to that paper. 



The figures may still be taken as correct representations of the 

 layers, but I am afraid this is the only part not rendered obsolete 

 by subsequent observations by myself, and most of the theories 

 proposed there must be thrust aside as crude deductions from 

 imperfect observations. The figures of P . formosiim may stand, 

 but that given of the middle layer is not the middle layer at all 



