004 MESSRS. W. HILL AND A. J. JUKES-BROWNE [Nov. 1 895, 



while the Cyrtoid test is only feebly outlined. Radiolarians with 

 the sponge-like character of the test can be sometimes recognized 

 by a greater opacity of the material within the outline. 



While the siliceous network of the test is frequently obliterated 

 in these Miocene radiolarians, one structural characteristic is some- 

 times retained, — namely, ' the medullary shell.' This is often defined 

 by an inner ring, a clearer central area, or a central area filled with 

 the denser semi-opaque material of the matrix of the rock. Even 

 when the whole test is completely filled with clear material, and no 

 trace of the skeleton is left, the position of the medullary chamber 

 can often be seen. Eadiating spines, the striking feature of many 

 radiolarians, are not well shown in sections of Barbados rocks. 



Thus it will be seen that, while in calcareous rocks where radio- 

 larians are indubitably present it frequently happens that nothing 

 but their outline remains, our slides show all stages in the trans- 

 formation of a siliceous radiolariau into a structureless ball or disc 

 filled with calcareous matter or into a mere patch of clear crystalline 

 material. "Without the aid of these slides we could hardly have 

 understood how the radiolaria of the Chalk have become so completely 

 obscured as we know them to be. 



It must be premised, therefore, that in the forms about to be 

 described the siliceous skeleton of the radiolarian is lost, and the 

 body now visible under the microscope is the section of a cast — 

 either in crystalline granular calcite or in fine calcareous paste. It 

 is this fact which has prevented the earlier identification of the 

 organisms, for all trace of the beautiful radiolarian structure is 

 gone, and nothing remains but a mere patch, which is, however, in 

 most cases easily distinguishable from the surrounding matrix. 

 Many of them have a clearly marked periphery, and others are 

 definite spheres of crystalline calcite, but in no case is there any 

 trace of the actual cell-wall. 1 



This absence of a test is a character which at once distinguishes 

 these radiolarian remains from almost all the other organic bodies 

 which occur in the Chalk, and especially from all kinds of forami- 

 nifera, the tests of which are always preserved. Had we not the 

 proof of the manner in which the radiolarian test disappears, it 

 might have been supposed that it would have been simply replaced 

 by calcite, as in the case of sponge-spicules, which are' frequently 

 so replaced, though in some parts of the Chalk Marl they remain 

 unaltered. 



IV. Description op the Porms met with. 



As the chief characteristic of the radiolarians of the Chalk is 

 wanting, the siliceous skeleton having disappeared, it is impossible 

 to refer the various forms we have met with to their respective 

 orders and genera with accuracy, but it is necessary to describe 

 them in some way. Therefore in provisionally arranging them we 



1 Since writing this paper, on examining a slide which had been overlooked, 

 we found a single specimen in which the network of the skeleton can be 

 faintly traced ; the silica is entirely replaced by calcite. This form, shown in 

 PI. XXII. fig. 16, may be referred to the family Liosphserida. 



