292 



KNOWLEDGE. 



AL'GI'ST. 1911. 



image of the whole. No increase of magnifying 



power will separate the la^-ers. then, though ont 



may know them to be there. 



The real cause of failure '^ 



may be explained in this way. 



Suppose a sheet of trans- 

 parent paper with writing 



upon each leaf, differing, 



perhaps, in both character and 



subject. Clearh'. if in close 



contact nothing could he read — 



separate the leaves to a proper 



distance, when both writing 



and meaning become plain. 



The PlL'iirosi}>nia valve, mounted 



in a medium, is the sheet of 



transparent paper with the 



leaves together: mounted dr\-, 



the same sheet with the lea\'es 



well separated. Figures 5 and 



6 will illustrate this point, 



though taken with different 



lenses. Figure 5 exhibits cer- 

 tain structure, discovered h\- 



the writer, upon the outer 



side of a valve, P. formosiim. 



mounted dr\-, as seen under 



an oil immersion. Figure (> is 



from the same vah'e, and the 



same iwint, taken with Zeiss" 



six millimetres aperture of -''5 



N..A.., a dr\' lens. The struc- 

 ture seen in the first }>rint is 



here totalh' in\isilile as a recognisable image, though 



there are certain indications in some scattered dots. 



of which no one could discover the reason if taken 



1)\- themselves, without the ke\'. In both instances 



the magnification is the same. 



Figures 7 and 8 are further exam[)les illustrating 



the same point. The 

 subject is what the 

 w riter has called his 

 " ru n n ing dnw ii 

 case." One minute 

 diatom vahe had 

 collided with aimther 

 and rijiped up the 

 ilcck. if one ma\ be 

 allowed to use such 

 a term. In Figure 7 

 the torn structure at 

 J the sides is seen 

 J hanging in strips: so 

 shallow is the focus 

 of the oil immer- 

 sion with which this 

 was taken. that there 

 are no indications 

 of structure under- 

 in the microscope the tine 

 It will be of interest to 



note here that the two \al\es [present opposite sides 



to the eve. Figure S of the same incident is taken 



with the same dry lens as No. 6, but tells no story : 



not due to insufficient resolving 



» . power, as we shall see in an- 



•••:': other print, onl\- that the 



■. . ' objective has too deep a focus 



to separate the layers, or even 



the \alves. 



We are in this jiosition. 

 then : we cannot work with 

 an oil immersion through suc- 

 cessi\-e lavers of structure when 

 there is air between — that is, 

 not effectively — or separate 

 tluni opticallv in a medium 

 when the focus is too deep. 

 This happens with all the 

 6[jecies of diatoms given as 

 examples here, except the last, 

 and there seems to be only 

 one other wa\'. 



We must examine spread 

 slides in a dr\' mount, and 

 note the difference in appear- 

 ance and curves between the 

 \al\es : stud\' them separately 



remembering that, under 



FicuRK S. Tlie saiiR- specimen as s( i n iii 

 Figure 7, X 1720, taUeii with the same lens 

 and under the same conditions as l-'igure 6. 



Figure 9. The outer side of 

 PI euros igm cj fonnos urn, showing 

 " chains " of fibrils giving rise to 

 the appearances seen in Figure 

 3, X 1750. 



neath, though of coursi' 

 adjustment would gi\e it. 



then. 



iin oil immersion, only those 

 tight against the cover will 

 give the distinctive image. 

 Referring back to Figures i 

 and 4. it will be seen that in 

 Figure 3 onl\- along the middle 

 is it in sharp focus, shelving down on the upper side 

 into shadow. Figure 6, however, shows the distinc- 

 ti\'e curve of the outside of the valve still better 

 (though not the distinctive structure) : on that side 

 it is \'-shai)ed in section. In I'igure 4 of the same 

 diatniii. the inside, it is found to be flat across, except 

 m the middle, where 

 that and the median 

 line are seen bv the 

 shadow to be below 

 t h e g e n e r a 1 1 e \' e 1 . 

 The same difference 

 of cur\'e between the 

 two sides of the \'alve 

 occurs also in P. an- 

 giilatiiin. and seems 

 always to accompany 

 the same distincti\'e 

 appearance. 



These details may 

 seem tri\ial. \'et are 

 necessarx in clear- 

 ing the wa\' to- 

 wards getting at the 

 With rei/ard to the 



FiGl'RE 10. A similar specimen 

 to that seen in Figure 9. 



ke\- to the real 

 outer side of P. 



Figure 9 seems to 



structure. 

 ioiDiosiini, 

 supph' the key to I'igure J. 

 The structure here is fairly sound, but in some 

 wa\- has been acted upon to leave it bare. It 

 appears to consist of a series of chains, as it were, 



