124 THE VOYAGE OF H.M.S. CHALLENGER. 



to the exterior surface of the plates a slightly convex aspect. The compound plates are 

 built up in the following manner : — from the upper surface of a more or less regularly 

 perforated simple plate, that is to say, from the innermost layer of the compound 

 plates, which thus seems to be developed first, a number of minute processes rise, from 

 the tops of which branches run out which join with one another and constitute an 

 irregular network, which in its turn gives rise to another net-work which lies above it, &c. 

 Oneirophanta furnishes most evident examples that the development takes place as above 

 described, because the plates, though they are simple, carry on their upper surface one 

 or several small processes (PL XXXI. fig. 1), which in most cases remain unbranched, 

 but which sometimes give off small branches from their tops, which unite and form a 

 rudimentary network on the upper surface of the plate. 



Both plates and wheels take their origin from the same typical form, viz., a small 

 spicule provided with four short arms (PI. XXXII. fig. 3), and more than once I have 

 had occasion to state the correctness of this view. In its first stage of development 

 the plate always presents the appearance of a spicule, the four arms of which increase 

 in size and give off branches, which, connected with each other, form the larger holes 

 which are always situated in the centre of the plate ; round the edge of this primary 

 plate new holes arise successively by means of the development of new processes 

 which become connected with one another. Though it may seem very strange, the 

 wheels in the Elasipoda are developed in the same manner. As has been already noted 

 the wheels in this order are remarkable for having a small central crown made up of 

 four, rarely five or six, short arched arms ; this crown, which ought to be regarded 

 as a four- to six-armed spicule (PI. XXXII. fig. 3), becomes first visible, aud its more 

 or less curved arms, being linked together by their ends so as to form four to six holes, 

 give rise to an irregular ring (PI. XXXII. fig. 2), which is the first indication of any 

 nave. This nave increases in size and gives off round the edge small, conical processes 

 (PI. XXXII. fig. 4), which growing larger become spokes which in their turn send out 

 branches towards each side ; these branches joined together with one another form the 

 felly. The wheels in Myriotrochus, Steenstrup, Trochoderma, Theel, Chirodota, Eschsch., 

 &c, which lack the crown as well as the large hole in the centre of the nave, are de- 

 veloped in a slightly different manner. In these forms the wheels do not take their 

 origin from a spicule but from a calcareous star, which gives off as many small processes 

 or rays as there are spokes ; the further development takes place as above described. 



When the calcareous bodies are examined with a high magnifying power and 

 treated with a dilute acid and some colouring matter, it will easily be seeii that the}'' 

 are surrounded by a thin membrane which sometimes seems to communicate with fine 

 threads, which either belong to the surrounding connective tissue or are nerve fibres. 

 A central canal is often discernible within the calcareous deposits. In the hope of 

 obtaining a favourable specimen for showing how the calcareous deposits are developed, 



