312 M. F. Dreyer on the 



majority of the single-shelled forms are perforate, and the 

 pylornatic-monaxonic forms are in the minority, although they 

 do not fall so far behind the others as does the perforate forrn- 

 type among the monothalamous Thalamophora. This dis- 

 tinction in the distribution of the two growth-types becomes 

 still more strongly marked, however, in the many-shelled 

 forms with secondary growth. Thus in the Eadiolaria both 

 growth-types occur widely distributed side by side, but still 

 in such a manner that a preponderance of the concentric 

 growth is unmistakable, while, on the contrary, in the 

 Thalamophora the terminal growth-type is exclusively* 

 represented. 



The cause of this different behaviour of the Thalamophora 

 and Eadiolaria is to be found in the fact that the two modes 

 of construction in question make different demands upon the 

 solidity of the material. The perforate-concentric shell-con- 

 struction requires much finer material than the pylomatic- 

 terminal, and therefore it happens that, while in the siliceous 

 skeletons of the Eadiolaria both shell- constructions ai*e 

 represented in the highest completeness and complication, 

 the Thalamophora are under the necessity of producing ex- 

 clusively pylomatic-terminal shells, for with their material of 

 construction, which is softer in comparison with silica, it 

 would not be possible for them without impairing the solidity 

 of their shells to form concentric and airy skeletons like those 

 of the Eadiolaria ; they must make their shells thicker and 

 more massive in order to give them the necessary solidity. 



It is in the essence of the perforate-concentric mode of con- 

 struction that it requires to be carried out more lightly. As 

 there is no principal orifice, the passage of the sarcode to the 

 outer world, and in many shell-forms also between the dif- 

 ferent interspaces of the shell, is consigned exclusively to 

 the pores of the shell, which for the purpose of ready- 

 communication must not be too narrow nor the intervening 

 skeletal parts too massive ; further, the union of the latticed 

 spheres concentrically nested one within the other is only 

 possible by means of free radial rods, which, again, must not 

 exceed a certain thickness. The conditions of the pylomatic- 

 terminal mode of construction are very different. Here the 



* Only one remarkable exception to this rule is furnished by Thuram- 

 tjiina papillata^ Brady, the agglutinated shell of \\hich is composed of two 

 concentric spherical shells united to each other by some radial beams 

 (Brady, ' Challenger ' Report, pi. xxxvi. fig. 12). The stout and rather 

 irregular character of this form shows us, however, that we have here to 

 do as it were only with an unsuccessful attempt to imitate the light 

 ponstructiou of the siliceous skeleton with a less solid material. 



