6 Dr. J. D. Macdonald on the Structure of the 



Fig. 3 illustrates the ultimate separation of the two new 

 frustules, in which the same process is repeated : A, the pri- 

 mary-valve frustule ; B, that of the secondary valve. The 

 smaller letters and numbers correspond with those in the fore- 

 going figures. 



In fig. 4, 1, 2, and 3 are ordinary examples of Biddidphia^ 

 drawn from nature for comparison with figs. 1, 2, 3 respec- 

 tively, which are diagrammatic. 



On submitting a large frustule of Isthmia to microscopic 

 examination, the pitting or markings of the invaginated hoop 

 may be distinctly focussed through the enclosing or external 

 one connected with the primary valve ; and it is remarkable 

 that the artists employed by various writers to illustrate their 

 works have shown this unequivocally in many instances, mi- 

 litating irreconcilably with the text. 



The " hoops " of the tertiary valves are gradually evolved 

 as the new frustules progress in development ; and even while 

 included within their formative valves their markings are often 

 clearly discernible. 



Before the tertiary " hoop " is of sufficient depth to give 

 the new frustule its adult dimensions, the outer " hoop " will 

 be seen to extend beyond the gibbous fulness of the younger 

 valve in Biddulfphia and Isthmia — a condition which is incom- 

 patible with the idea of a single " intermediate piece " or 

 " connecting membrane " of the existing theory. Any deep- 

 ening of the so-called connecting membrane is therefore only 

 likely to happen in connexion with the tertiary hoop, no addi- 

 tion being necessary, nor, indeed, at all capable of proof, as 

 respects the adult or outer " hoop." Dr. Wallich, however, 

 assumes that additions are made to both in the manner above 

 alluded to. 



Of course, in particular genera, where the hoops of the 

 valves are exquisitely thin and destitute of markings, it would 

 be more difficult to trace out the particulars just described. 

 The inference, however, appears legitimate, unless sufficient 

 reasons can be advanced to warrant a contrary opinion, that, 

 small and large, the same general laws of development obtain 

 with all the Diatomacete. To conclude these remarks, it must 

 be stated that every tertiary valve becomes in due course 

 secondary, and ultimately primary, beyond which there is no 

 further advance ; but after having been the parent of an almost 

 unlimited progeny it must tend to decay, if it be not privileged 

 to close a genetic cycle by taking part in the development of 

 a sporangial frustule from which another living chain may 

 descend with renewed energy. 



Mr. Kalfs uses the term " front view " for what appears to 



