CASTRACANE, ON DIATOMACE^. 261 



This colour, from the bright green hue of chlorophyll, passes 

 into a glaucous or bluish-green, olive-green, and yellow, 

 until it assumes the rusty yellow or ochraceous tint belonging 

 to the endrochrome of the perfect or adult diatom. This 

 observation of mine accords with a circumstance noticed by 

 Mr. O'Meara in Pleurosigma Spencerii, which at the moment 

 of emitting the germs exhibited a green colour, which, on 

 the following day, had become olivaceous. This seems to me 

 confirmatory of the view that the endochrome of the 

 Piatomaceae is composed of chlorophyll, which takes on the 

 ferrugineous yellow or ochraceous colour in proportion as it 

 assimilates iron, the presence of which metal in the Dia- 

 tomaceac has been proved by the analyses conducted by Pro- 

 fessor Frankland at Manchester. And the identity thus 

 proved of the endochrome of the diatoms with chlorophyll 

 affords a further insuperable argument in favour of their 



rotable nature. 



After these observations I was further desirous of subject- 

 ing to the action of nitric acid some of the green masses in 

 the aquarium above mentioned, and which I judged to contain 

 nascent diatoms, with the view of proving the presence of 

 silica in them, and possibly of determining the period at 

 which that mineral element is developed. I conducted the 

 experiment with the utmost care I could bestow, so as, in 

 the repeated necessary washings, I might lose as little as 

 possible of these delicate corpuscles. From the minute 

 traces of siliceous matter thus procured as the ultimate pro- 

 duct I mounted a preparation in Canada balsam ; and 

 although the embryonal forms had been inevitably lost, I was 

 able ( le irly to distinguish, though unusually small, Nitzschia 

 minuti&sima, linearis, and amphioxys, Pinnularia radians, 

 and an Amphora. l>ut in order to discern these I was 

 obliged to employ an oblique illumination, to which v, 

 adapt* d an excellent objective No. 10, with correction for 

 immersion, by Hartnach. [n the same preparation, besides 

 oth difficultly recognisable forms, were some of extreme 



minuteness, in which I was unable to distinguish any d< bails 

 on the surface of the valvesj and others, again, which 1 v. 

 able to det< rmine, are of such astounding minuter I 



have hitherto never witnessed in all the numerous circum- 

 stance under which I i studied these Bpeci 



This would be the place to consider the question whether 

 the frustule, when once fornn d, is capable 01 further develop- 

 ment ox growth, and if new striae continue to be added to 

 the \ dve- ; or it. on the other hand, those already existing 

 ma] 'me wider apart, so that in .1 given space of the 



