ANIMAL NATURE OF DIATOME^. 467 



(rectangula, obtusangula} of the four species of this genus 

 a continuation of the vittse from one margin to the other, 

 instead of their being interrupted and alternating, as they 

 are figured and described by Kiitzing. Continuing my 

 observations, I succeeded at last in finding one individual 

 exhibiting to my sight the alternations described. Hence 

 I became convinced that the latter condition is not merely 

 inconstant but even the least frequent. The secondary 

 surfaces, neither described nor figured by Kiitzing, are 

 elliptico-acute, and on these are inscribed the smaller 

 concentric ellipses, which mark the margin of the in- 

 complete diaphragm formed by the nearest vittae. 



53. RHABDONEMA. Bacilli tabulati concatenati late- 

 raliter stipitati, interrupte vittati et transversaliter 

 striati; vitta capitatce. Stria transverse sinis longitudi~ 

 notes numerosas formantes. 



I have only had an opportunity of observing one (R. 

 adriaticum] of the three species of this genus ; and it is 

 because I must therefore limit myself to this that I can- 

 not agree to what Kiitzing represents in his description 

 and figure. He omits to indicate the form of the tablets, 

 which can only be obtained by observation of the 

 secondary surfaces, or by a longitudinal section ; and it 

 is entirely from this form that the appearances are derived, 

 on which the specific distinctions are supported. In the 

 Adriatic species, the figure of the secondary surfaces is 

 linear in the centre, and cuneato-attenuated (assottigliata} 

 at the extremities. The frustules (bacilli}, therefore, are 

 thick in the middle, have laterally the two primary 

 surfaces strongly inclined together towards the outside, 

 and are much attenuated in the extreme margin. The 

 vittae are nothing more than canals projecting into the in- 

 ternal cavity, and projecting slightly, indeed, at the sur- 

 face, in the middle, and continuous from one extremity to 

 the other. In specimens dried and softened again, 

 the canals themselves include air collected irregularly 

 into bubbles more or less extended. Hence the appear- 



