312 



Dr. H. H. Dixon. On the Structure of 



slimy layer, exhibits a slight orange coloration when treated with 

 nitric acid followed by ammonia. With Schultze's solution it becomes 

 amber-coloured. 



Within this internal membrane in all my specimens but scanty proto- 

 plasmic contents were revealed, but in these there was often distin- 

 guishable a minute, more darkly staining body, presumably a nucleus. 

 In several specimens this nucleus was double or hour-glass shaped 

 (figs. 12 and 15). My specimens did not show the structure of the 

 nucleus with precision, and nothing beyond its granular appearance 

 and probable possession of a membrane could be made out. The aggre- 

 gation of protoplasm in which the nucleus is situated occupies a 

 lateral position, and is in contact with the internal membrane, and 

 strands of protoplasm extend from it across the cavity of the sphere. 

 In no case was a chromatophore, nor anything like one, seen in the 

 protoplasm. In one case a colourless trilobed body was found in the 

 cell ; its nature is quite uncertain. Fig. 16 is drawn of it after its 

 containing coccosphere was treated with liquor iodi. The unfavourable 

 conditions of our coasts are, perhaps, responsible for the scantiness of 

 the protoplasmic contents of the coccospheres obtained in the south of 

 Ireland. But to these unfavourable conditions can scarcely be attri- 

 buted the absence of the chromatophore, for the existence of which in 

 C. pelagica we have no definite evidence. 



I now go on to a series of observations which seem to me to be of 

 considerable interest, as throwing some light on the manner of growth 

 of coccospheres, and on the origin and development of coccoliths. If 

 entire coccospheres are examined in a medium of high refractive index, 

 e.g., balsam, or even glycerine jelly, it will be found that numerically 

 about 80 per cent, of them contain an internal oval colourless body. 

 Closer examination reveals that this body is in many cases a complete 

 and perfect coccolith (fig. 17), in others it is a simple oval ring or 

 shallow collar (figs. 3, 7, 8, 15). All stages of development connecting 

 the collar-shaped body with the complete coccolith are found, so that 

 it becomes evident that the coccolith arises as a ring of calcium car- 

 bonate within the coccosphere. At first the ring is a narrow band 

 (fig. 3), it then deepens into a collar (fig. 8) ; on either end of the 

 collar are then secreted oblique flanges (fig, 15), which, as deposition 

 continues, are developed into the bevelled valves of the coccolith 

 (figs. 7 and 17). The central body appears last, and only in its later 

 stages is the transverse bar secreted, dividing its single aperture into 

 two. The position of the internal coccolith within the coccosphere 

 varies ; generally speaking, in its earlier stages of development, it 

 lies near the centre of the sphere, and in many cases it was found in 

 contact with the nucleus (figs. 15, 17). When more mature it comes 

 into contact with the inner gelatinous membrane, and when the cocco- 

 sphere is intact, appears in close proximity to the external coccoliths 



