ZOOLOGY AND BOTANY, MIOKOSOOPY, ETC. 893 



connectives. Apart from these questions of detail, the concordance is 

 complete.* 



In order to elucidate the nature of the striae or ribs of Pinnularia, 

 I have examined sections parallel to the long axis of the diatom, thus 

 intersecting perpendicularly the ribs which run along the valves. 

 These longitudinal sections are present in great numbers in the thin 

 laminae of Franzensbad earth ; they are rarely defective. When such 

 a section is too thick, it includes a more or less considerable portion 

 of the raphidian region or a portion of the edge of the valve. Sections 

 of a thickness equal to c a, c e, a d, h e, fig. 155, would appear thus : if 

 we look at a section of the thickness of c a, for instance, with a pene- 

 trating objective, the ribs invariably appear to be closed in their upper 

 part by a membrane ; in fact, they resemble sections of cylinders. 

 This image proceeds from the fact that the objective shows at the same 

 time the section of the ribs, and also the lines which represent the 

 deeper portions of the valve.f The result is a 

 single image, difficult to interpret otherwise jij^ ^gg 



than by the presence of cylinders, fig. 156 (1). 



But if we use an objective without penetra- s CTr") n r-»rS 7=^A~~ 

 tion, it corrects, in some degree, the too great sWUtJUUCjCl 

 thickness of the section, by separating the differ- 

 ent planes which constitute it. The images, 2 

 instead of being blended, succeed one another, (\ [\ f] (] f] n n n 

 and the cause of error is removed, fig. 166 (2). £^ UUUUUUt^ 



We have thus far considered rather thick 

 sections. In those which are thinner, the edge 

 of the valve and the solid portions which run JlJlAilAAA/l/l 

 along the raphe are removed, and there remains i ..i 



nothing but the section of the ribs themselves t^,. i--^ /ix • 



/7 ■, ry iKKN Fig. 156(1) IS seen 



(& cZ fig. 155). _ with a Zeiss diy 1.14th 



An image is then obtained, fig. 156 (3), in. (2) represents the 



which can leave no doubt as to the presence of s«"i« section seen with 



parallel elevations separated by spaces rather ? ^^^^^^ homogeneous 



•J J.I, J.I, • XT-' 1 ^ J. immersion 1-lOth in. 



Wider than their thickness. ^^s ^^^^ not chan 



I shall not insist upon the utility of sec- whatever the objective 

 tions for the study of diatoms ; it is by this employed, 

 means only that we can hope to have exact 



notions of their structure. AH the methods for obtaining them are 

 not equally good. Sections which are prepared by inclosing the 

 frustule in gum may give rise to false interpretations. This medium 

 not being perfectly solid, the sections are liable to break up in their 

 most delicate parts. Suppose for a moment that the raphe is only a 



* Dr. Flogel has also studied sections of Pinnularia. His conclusions are 

 only known to me in a very short resume in the Bot. Ztg., 1872, p. 471. 

 The skilful micrographer there describes some types of diatoms seen in section. 

 Pinnularia shows in section, according to him, "a large empty space, which has 

 only one passage, in the form of a short canal, towards the interior." 



t The drawing by Walker Arnott, reproduced by Mr. Burgess (loc. cit. 

 fig. 27), represents a fragment of a valve equal in thickness to 6 c of fig. 155. 

 Perhaps it includes the whole thickness of a half valve a e. 



