Dr. 0. Schmidt on Coccoliths and RhabdoUths. 365 



queiit division of the margin of the medullar space. The 

 margin of the medullar space swells up ; and such specimens 

 as fig. 5 show that the marginal swelling is not formed 

 uniformly, and only by degrees constitutes a perfectly closed 

 circlet. Almost without exception, the whole structure, at 

 the appearance of the circlet of balls, acquires the form of a 

 bowl, at the same time that there is the greatest variability in 

 the number of the globular portions and the size of the individual 

 globules. 



In two cases we have nothing more formed than a mere 

 zone of globules or granules. This is shown in figs. 6, 7, 9, 

 10, 11. One case is, that only a circle of large globular por- 

 tions surrounds the medullar ring. A very frequent case of 

 this kind is shown by fig. 6, from the fresh Batliyhius^ and 

 fig. 9, from the raised deposits near Brindisi. It is true, as 

 I shall show hereafter, that here also completion may be 

 effected by the outer ring, i. e. the dorsal disk ; but in general 

 the development ofthe body seems to be closed with the growth 

 of the large globular portions ; and I have no indication that 

 these large globules break up into smaller portions. Such bodies 

 as fig. 11 are excessively rare. It shows, in the circumference 

 of the central disk, a circlet of large celliform balls ; and this 

 body also occurs but little altered in the deposits at Brindisi 

 (fig. 10). The centres in the globules of the circlet appear 

 quite distinctly to be condensed into a sort of nucleus, which 

 also makes its appearance quite definitely, although irregularly, 

 in the specimen from Brindisi. 



In the second case (fig. 7) we find, instead of a simple 

 circlet of globules, a granular zone, i. e. a zone in Avhich small 

 globules of about the diameter of the central granule are placed 

 in from 2 to 4 rows side by side and partly over one another. 

 Both this and the above described body may, as we show 

 further on, cover themselves with the dorsal shield ; but fre- 

 quently this course of further development does not occur, and 

 bodies are produced like fig. 8 and, more distinctly, fig. 12; 

 they are hemispherical. In fig. 12 we are looking into the 

 sphere ; fig. 13 is the appearance from the side, fig. 14 from 

 behind. Instead of a granular zone, a regular granular mantle 

 is present. I am uncertain about the observation that some- 

 times an entire hollow sphere is produced. How the division 

 into globular portions is compatible with the not unfrequent 

 concentric striation, is not yet clear to me (fig. 8). 



I give on PI. XVII. figs. 16 & 17 a form of the incomplete 

 coccolith which is not unfrequent. The central granule and 

 medullar space are present, the latter imdoubtedly as a disk. 

 There is, however, no medullar ring ; but what might be 



