THE CONJUGATION OF THE DIATOM EHABDONEMA ARCUATUM. 135 



it, and one of the isthmi connecting the female with the frustule 

 below it gives way, leaving it cohering by one angle. 



Later still the males lose all their green colour, and the contents 

 are still further contracted, so that only amorphous minute specks 

 are left that look like the denser remnants of the original granules, 

 and the frustules break in two, — either near the middle, or lose one 

 valve: having fulfilled their office their vitality is quite lost. The 

 stimulus they have exerted upon the contents of the female cell 

 has passed through the thin (possibly unsilicified) spot near the end 

 of the valve, and has undoubtedly been conveyed by the isthmi 

 which connect them with the annuli of the female, and, I think, 

 must have made its way through the sutures of these annuli. The 

 endochrome of the males appears mainly to pass into solution to 

 effect the stimulus : there is obviously no transfer of the granules 

 as such. The female has developed — projecting from the band — a 

 globose gelatinous sporangium, in which there is a distinct small 

 portion that contains most of the endochrome. (Fig. 6, X 250.) 



Succeeding this the whole of the endochrome passes into this 

 space, and then appears completely separated from the cavity of the 

 frustule, and is gradually elongated, and has formed round it a 

 sheath or membrane. 



Finally, the elongated mass secretes silex, and a new frustule is 

 formed. The gelatinous investment has grown with the require- 

 ments of the contained body. The new frustule consists of two 

 valves and what seems a kind of hoop, — there are no annuli, — and 

 it contains larger, denser, and more deeply-coloured granules. 

 Some of the male frustules still hang attached although it is evi- 

 dent, by the still further diminution of their opaque contents, that 

 it is only by the persistence of their isthmi. (See. Fig. 7, x 300.) 



With regard to the double sporangia a substantially similar course 

 of development is followed. In some cases, however, one sporan- 

 gium appears to be more advanced than the other. In Plate VIII., 

 Fig. I,x200, the lower one was either started earlier, or has de- 

 veloped more rapidly. In this specimen the female frustule was 

 not terminal — a frustule being beyond it. In Fig. 2 — showing two 

 mature sporangial frustules ( x 300) — it was terminal. It is now- 

 rare to find the female frustule of the upper sporangium. Very 

 frequently it gets rubbed off, even by the time the stage in Fig. 1 

 (preceding) has been reached, and only fragments of the band are 

 left, but this is probably of no consequence, the developing poten- 



