G60 SUMMARY OF CURRENT RESEARCHES RELATING TO 



but largo processes, from which run to the centre conspicuous double 

 lines of largo areolae, giving to tbo valve the well-known appearance 

 of being marked with a cross, tbc ends of which end in the hollow 

 side of the crescent-shaped processes. In the two valves these are 

 not opposite each other, but alternate ; and so long as two or more 

 frustules remain in a temporary filament, they are interlocked by 

 each process fitting into the hollow between those of the neighbouring 

 valve. The hyalino hoops arc divided by well-defined sutures into 

 five, six, or more parallel bands or rings. Tho corrugated appear- 

 ance of the hoop is an optical illusion, as is shown by the fact that 

 the hoops slide over each other when the new frustules separate after 

 the self-division of the parent. The most noticeable characteristic of 

 these hoops is that the sutures are not continuous lines going quite 

 round the shell ; but at one place they curve sharply upward towards 

 the valve, so that a tooth from the next outer division of the com- 

 pound hoop cuts through its neighbour. These teeth alternate upon 

 different sides of the shell ; and if the hoop be divided upon the lines 

 of these sutures, it will be found to be made up of a connected series 

 of imperfect rings or bands with a projecting tooth upon the edge, 

 and with the curved ends of the band separated by a space into which 

 would fit a similar tooth upon the adjacent band of the hoop. The 

 direction of the teeth has a fixed relation to the valve to which each 

 hoop is attached, uniformly pointing towards it. 



When the fission of the parent diatom is complete, and the two 

 new frustules slide apart, the hoops have ripened so that the sutures 

 between the bands open at the slightest touch. The division of each 

 band by the tooth of the neighbouring one allows it to spring open, 

 and the frustule (or pair of frustules ready for separation) is thus 

 freed from the hoops, which fall to pieces of themselves. This 

 mechanism Dr. Cox believes to be designed to facilitate the escape of 

 the new diatoms from the shell of the old one. Similar structures 

 are found in other genera of the family. 



Division of Stephanodiscus Niagarse.* — From observations of 

 this diatom, Mr. C. M. Vorce confirms the conclusions of Dr. J. D. Cox j 

 with regard to the hoop. By examining continuous series of gatherings 

 the entire process of division may be seen. The first change observed 

 is the widening of the connecting zone or hoop. The box of tho 

 frustule becomes in consequence deeper, until it is often as deep as 

 it is wide ; and at the same time the endochrome increases in quantity. 

 About this time there is seen an extremely fine line of division 

 crossing the centre of the frustule in the middle of the central mass 

 of endochrome, almost invisible at first, but gradually becoming more 

 distinct, and at first soft and flexible. Later it becomes doubly clear 

 across the frustule, and begins to exhibit indications of the future 

 spines of the new valves. The frustule has now become double, 

 composed of two frustules, each of which has its outer valve thick 

 and strong with long spines, and its inner new valve thin and fragile, 



* Proc. Amer. Soc. Micr., 8th Ann. Meeting, 1885, pp. 139-41 (4 figs.), 

 t See preceding notice. 



