478 ANIMAL NATURE OF DIATOME^. 



that in A. undulatus the six apparent rays are produced 

 by a similar flexuosity, a doubt necessarily arises, either 

 that the two genera are not sufficiently distinct, or that the 

 characters deduced from this flexuosity are not sufficient 

 to distinguish the species. There occurs, also, naturally 

 to the mind, a comparison under this aspect of the 

 Coscinodisci with Campylodisci and the flexuose SurireU(e. 



In some individuals exactly referable to Coscinodiscus 

 eccentricics, and which preserves both the valves, there 

 are seen five incomplete radii, which vanish entirely when 

 the object is withdrawn so far from the microscope that 

 the surface only is visible. I conclude from this that 

 these radii are so many internal partitions (setii). On 

 this supposition, the species would not belong to any one 

 of the three genera. Besides Ehrenberg having also 

 found it alive near Cuxhaven, in the North Sea, it would 

 seem improbable that such an organic condition could 

 have escaped so expert an observer ; and we must, there- 

 fore, suppose that a different thing is referred to. 



Kiitzing describes and represents the radii to be per- 

 fectly smooth ; Ehrenberg, again, though he does not 

 express it in his figures, says that they are finely punctated. 

 In almost all the species there are indicated marginal 

 apertures corresponding to the radii. In many, which 

 Ehrenberg observed alive, the coloured internal substance 

 was variously grouped into many lobes near the centre. 



60. AcTiNOPTTCHUS. — Individua soUtaria libera, lorica 

 bivalvis silicea disciformis (breviter cylindrica), ceUidosci; 

 cellidcB radiis septisque internis radiantibus pluribus 

 inierniptcB. 



The beautiful observations of Ehrenberg are fruitful in 

 very important particulars relative to some species of this 

 genus, {duodenariiis, sedenarius, ododenarius.) Of these 

 particulars, Kiitzing takes no notice. The triangles into 

 which the disc is divided by the radii, appear alternately 

 clear and dark, precisely as happens in Adinoci/clus 

 widulatus, where that appearance is evidently produced 



