1895 THE MICROSCOPE. 51 



dent may still be at a loss as to precisely what is meant 

 by this or that character. And so it has seemed to me 

 that it would be helpful to illustrate the key by some dia- 

 grammatic figures which would bring before the eye 

 simply the principal characters which occur in the course 

 of the classification. The diagrams are very rude but 

 they will answer the purpose. They are not intended to 

 represent any particular species but merely to give a gen- 

 eral idea of the structure in each case. 



I shall assume then that the reader knows nothing 

 whatever about the classification of the Radiolaria but 

 is simply acquainted with their general appearance as 

 he has seen them on a slide of Barbadoes earth in bal- 

 sam. Let us suppose him to have such a slide before him 

 and to be able to identify the forms as far as the genera 

 are concerned. How shall he go to work.^ 



The first thing to do is to find out whether the form 

 which is under observation is spherical (or approxima- 

 tely so) and has no single large opening, or whether it 

 has the shape of an egg, cone, or bee-hive with a large 

 opening at one end. Let us suppose that the former is 

 the case. Now that puts it in the Legion Spumellaria 

 and limits him to that part of the Key. Look again 

 therefore, and see if the shell is a true sphere, or an 

 elongated sphere (prune-shaped), or a flattened sphere 

 (disk-like). We will say it is a true sphere. Then it 

 belongs to the order Sphseroidea. What now about 

 spines radiating from the surface ? Is the form free 

 from them like fig. 1 ? That places it in the Family Lio- 

 sphserida. Now is the sphere hollow? It belongs to 

 division A. Put on an ith objective and if the surface 

 is smooth it is Cenosphsera ; but if the pores project in 

 the form of little tubes just a trifle beyond the margin 

 it is Ethmosphsera. But perhaps as you focus down on 

 your sphere you find it is not hollow but that there is 

 another sphere inside, like fig. 2. Then it is div. B, Car- 



