722 ON CONCRETIONS, SPICULES, ETC. [ch. 



at 120°; and faces and edges become curved, whenever necessary, 

 in order that these conditions may be fulfilled. The need may arise 

 in various ways. Suppose (as in Fig. 339) that our httle central 



bubble be no longer in the centre of 

 symmetry, but near one corner of the 

 enclosing tetrahedron; the short edge 

 running out from that corner will tend 

 to remain straight (and so form the 

 spike of the helmet); while the other 

 three will each form an S-shaped curve, 

 as a condition of making co-equal angles 

 at their two extremities. An analogous 

 case is figured in one of Sir David 

 Brewster's papers where he repeats and 

 amplifies some experiments of Plateau's*. 

 Fig. 339. Diagram of one of the He made a tetrahedral cage, and fitted 

 helmet-shaped radiolaria, e.g. it with three more wires, leading from 



Podocyrtis: to shew its tetra- ,^ , ,-, • t t, • , n i i i 



hedral symmetry. ^^^ ^P®^ ^^ the middle pomt of each basal 



edge. On dipping this into soap-solution, 

 various comphcations were seen. At the apex, six films must not 

 meet together, for no more than three surfaces may meet in an edge ; 

 intermediate or interstitial films make their appearance, with which 

 we are not greatly concerned. But six films now ascend towards the 

 apex from the base instead of three, and the three which come from 

 the corners have a longer path than the other three which come from 

 the mid-points of the basal edges ; they must be curved in different 

 degrees in order that all t&ree may make at either end their co-equal 

 angles. And if we now introduce a bubble (or two bubbles) into the 

 interior of the system, we obtain the characteristic form of our helmet- 

 shaped Radiolarian — the spike above, the single or double vesicle 

 of the body, and the straps or lappets with their peculiar and 

 characteristic curvatures. 



The Httle shell is perforated with many rounded holes, and it 

 remains to account for these. They are required, so we are told, 

 for the passage of pseudopodia; well and good. We have referred 

 the Dictyocha-si^icule and the hexagonal meshwork of Aulonia to 

 froth-like associations of vesicles and to adsorption taking place 

 * On the figures of equihbrium in liquid films, Trans. R.S.E. xxiv, 1866. 



