DEPTHS IN THE NORTH ATLANTIC OCEAN. 211 



while, on the supposition tliat the coccospheres dehisce, it 

 must be supposed that the granular layer appears after de- 

 hiscence ; and, on both hypotheses, the fact that both cocco- 

 spheres and cyatholiths are found of very various sizes 

 proves that the assumed coalescence or dehiscence must take 

 place at all periods of development, and is not to be regarded 

 as the final developmental act of either coccosphere or 

 cyatholith. 



And, finally, there is a third possibility — that the diifer- 

 ences between the components of the coccospheres and the 

 cyatholiths are permanent, and that the coccospheres are 

 from the first independent structures, comparable to the 

 wheel-like spicula associated in the Avail of the " seeds " of 

 SpongiUa, and perhaps enclosing a mass of protoplasm 

 destined for reproductive purposes. 



In addition to Bathybius and its associated discoliths, 

 cyatholiths, and coccospheres, the Atlantic mud contains — 



a. Masses of protoi:)lasm surrounded by a thick but incom- 

 plete cyst, apparently of a membranous or but little calcified 

 consistence, and resembling minute Gromiae. It is possible 

 that these are unfinished single chambers of Glohigerinm. 



b. Globigerinoe of all sizes and ages, from a single chamber 

 -•s'-jpfith of an inch in diameter, upwards. I may mention in- 

 cidentally that very careful examination of the Avails of the 

 youngest forms of Globigerina with the Vnth leads me to 

 AA'ithdraw the doubt I formerly expressed as to their per- 

 foration. 



In the absence of any apparent reproductive process in 

 Globigerinoe, is it possible that these may simply be, as it 

 Avere, offsets, provided with a shell, of some such simple form 

 of life as Bathybius, Avhich multiplies only in its naked 

 form ? 



c. Masses of protoplasm enclosed in a thin membrane. 



d. A very fcAv Foraminifera of other genera than Globi- 

 gerina. 



e. Radiolaria in considerable numbers. 



/. Numerous Coscinodisci and a foAV other Diatoms. 



g. Numerous very minute fragments of inorganic matter. 



The Radiolaria and Diatoms are unquestionably derived 

 from the surface of the sea ; and in specu.lating upon the 

 conditions of existence of Bathybius and Globigerina, these 

 sources of supply must not be overlooked. 



With the more complete vicAv of the structure of the 

 cyatholiths and discoliths which I had obtained, I turned to 



