518 Transactions of the Society. 



tbat an able observer would take up tbe matter and confirm my 

 work, wbereby I should have been relieved of the trouble of 

 replying. In this hope I have for ten years been disappointed, 

 and I am obliged now to refute Miiller. 



I do not know whether Miiller has made Fleurosigma sections 

 according to my method, or whether his statements are made from 

 examination of my own preparations. He says that my drawings 

 are incorrect, especially that the diameter of the transverse section 

 of the walls is proportionately much too thick, whilst the strong 

 refractive thickenings at the ends of the same are not sufficiently 

 given, hence he will not admit the existence of closed chambers 

 (15, p. 621). 



With regard to this question, I have to reply that all my 

 preparations oi Fleurosigma were not only submitted to the late 

 Max Schultze, and every doubtful and difficult point demonstrated 

 before him by me personally, but that my diagrams were recognized 

 by him as correct, and by his express desire my paper was com- 

 municated to his ' Archiv.' This I mention without putting high 

 value on the influence of mere authority. Next I refer to my own 

 paper : on pp. 82-84 I discuss in considerable detail, with reference 

 to the coarsely marked P. halticum, the point as to the existence 

 of columns between two envelopes or closed chambers. Nobody 

 will infer from my description and diagram that I meant cylindrical 

 columns or chamber-walls of equal thickness, nor that I intended 

 to deny the thickenings at the ends. No such idea was in my 

 mind. But if such end-thickenings do exist it is an understood 

 thing that they are of pyramidal shape, the base towards the 

 membrane, the point towards the space between the membranes, 

 and they must operate as strong refracting bodies just like small 

 convex lenses. On p. 511 I illustrate for this purpose the most 

 striking comparison — the liver-wort leaf with large mesh-work. 

 Choosing for study sections of a considerable thickness, and in 

 which therefore two entire chambers might be found lying one 

 over the other, then the effect is doubled ; the two membranes 

 will be seen more conspicuously projected from the inner space with 

 the thinner walls. But such sections I did not select for my dia- 

 grams, it being the rule, whenever the finest structural details were 

 under investigation, to examine and to draw the thinnest sections 

 or the extreme margin as the most reliable portion. In examining 

 such sections one sees the detail exactly as I have represented it, 

 and I must continue to assert that my diagram is true to nature. 



A second point of attack to be disposed of is Miiller's idea that 

 his flooding experiments (14, p. 75, and 15, p. 621) could not be 

 brought into harmony with chambers closed from outside. I put 

 entirely aside the value of such flooding for the elucidation of 

 details of diatom structure. That all the fluids named by him 



