PRINCIPLES OF CLASSIFICATION. 41 



seems the more reason for including Gromida and even Lieherkiihnia within its limits, 

 when it is borne in mind that in the limitation of the origin of the pseudopodia to one 

 part of the body these forms bear a closer relationship to the Foraminifcra of the MUioUne 

 series, than the latter do to those of the liotalinc, in which the pseudopodia seem to extend 

 themselves equally from any part of the sarcode-body. 



52. Before enlarging upon the value of the differential characters just alluded to — 

 which will be shown to have such an important relation to certain peculiarities in the struc- 

 ture of the Shell, as to justify the employment of these as characteristics of the fundamental 

 divisions of the group — we must stop to inquire how far the separation of the Monothalaiiious 

 or Unilocular Foraminifcra, as an order distinct from the PoI^fJialamous or Multilocular (a 

 separation which has been adopted by D'Orbigny, Schultze, and Bronn), is to be regarded as 

 based on a right appreciation of their mutual affinities. It has been seen that, in common with 

 all the lower forms of animal and vegetable life, the Rhizopoda tend to multiply by a separa- 

 tion of continuously growing parts of their bodies, which may take the form either oi fssioii 

 or of gemmation, according as the original bod)'^ undergoes subdivision, or as it puts forth an 

 extension which eventually detaches itself. Among the Foraminifcra proper, whose bodies 

 are enclosed in unyielding shells, multiplication by fission cannot take place, except in that 

 early stage of existence in which the shell is not as yet consolidated (5[ 50) ; but extension 

 by gemmation may go on without limit. The progressive growth of the sarcode-substance 

 causes a portion of it to project beyond the aperture of the shell ; and this projecting portion 

 possesses all the attributes of the body of which it is an extension, and can maintain its exist- 

 ence with equal readiness either in a separate state (1[ 39) or in continuity with the stock of 

 which it is an offset. Although, therefore, there are certain types of Foraminifcra in which 

 such offsets appear invariably to separate themselves before the consolidation of their shell, so 

 that the original body never adds to the number of its segments and the shell remains 

 " monothalamous," — whilst there are others in which they ordinarily remain in connexion with 

 the original stock, so as progressively to augment the number of its segments and of the 

 chambers of its " polythalamous shell," often to an indefinite extent, — there is no such essential 

 difference between the physiological conditions of the newly formed segment in the two cases, 

 as would be required to justify the erection of the Monothalamia into a distinct order.* More- 

 over, we find that between these two groups there are gradational affinities of such a kind as 

 to render it impossible to separate them by a decided line of demarcation. For, on the one 

 hand, there are certain Monothalamous Foraminifcra which may be regarded as potentially 

 Polythalamous, the body and shell having the power of indefinite extension, but not exhibiting 

 any distinct segmentation; as is the case with Cornuspira and Spinllina, of which the former 

 is intimately related to the least speciaHsed forms of Miliola {% 104), whilst the latter is 

 scarcely less closely related to certain Rotalia. On the other hand, since there are certain 

 Polythalamia (f 40), the successive chambers of whose shells are so slightly connected 



* No botanist would think of separating from their natural allies, and ranking together as a dis- 

 tinct order, those Plants which habitually propagate themselves by detached gemnue, such as Lunularia 

 vulgaris and Lemiia gibba (which are only known to reproduce themselves after this fashion), or 

 Dentaria bidbifera, Gtobba amarantina, and Lilimn bulbifenim. 



6 



