Diatomaceous Frustule, and its Genetic Cycle. 7 



me to be in reality the " side view" of the frustule, corre- 

 sponding with the position in which the axis of growth is 

 perpendicular to the axis of vision, or, in other words, wliere 

 the component frustules would form a band or filament seen 

 sideways. On the other hand, when such a filament is seen 

 end on, the axis of growth being coincident with the axis of 

 vision, he would call the presenting end of the nearest frustule 

 a " side view," though this is unquestionably an end view by 

 all analogy. 



It appears to me that the axis of growth should be the 

 longitudinal axis, however short that may be, though the 

 broadest diameter is generally recognized as the length of the 

 frustule. 



Besides the siliceous envelope and the amber tint of the 

 contents, the Diatomaceee differ very materially from tlie 

 Desmidiaceaj in their process of self-division. Thus, in the 

 latter, more especially the solitary species, fission is attended 

 with the immediate and total separation of the valves, followed 

 by genuine gemmation of a new valve from each parent. In 

 the DiatomacefB, on the contrary, fission takes place under 

 cover of the sliding hoops, which retain the original valves in 

 contact, while the new endogenously developed half-frustules 

 are rather being modelled out of the preexisting material than 

 produced by genuine gemmation as in Desmidiace^. 



In the annexed diagram (fig. 5) I have attempted to trace 

 the history of a single Diatomaceous frustule through several 

 stages of self-division, expressing by simple figures the rela- 

 tive sizes of all the half-frustules. As a guide to the method 

 adopted, it will only be necessary to bear in mind that valve 

 No. 1 is the parent of valve No. 2, valve No. 2 the parent of 

 valve No. 3,' &c., the rising numbers representing the grades 

 of diminution — which are certainly not in a geometrical ratio, 

 like the simple multiplication of the frustules themselves. It 

 will be seen that all the frustules to the left of the median line 

 are the progeny of the primary valve, and those to the right 

 the descendants of the secondary valve ; and taking the grades 

 in their order, we find, on the primary side, 1 in the first 

 place, 2 in the second, 3 in the third, 4 in the sixth, 5 in 

 the eleventh, and 6, the highest number, in the twenty- 

 second. The highest numbers are also to be fomid in cor- 

 responding places on the secondary side, and the ratio is 

 certainly much more complex than the geometrical. 



Fig. 6 shows an undulating line by construction, giving 

 each valve of the long series its relative breadth within seven 

 thicknesses of the siliceous coat, but only representing one 

 border ; the progeny of valve 7 must exhibit a very ap- 



