148 Prof. W. J. SoUas on the 



The glohates agree in tlieir general character and mode of 

 development with those of other Geodina ; and I have now 

 only to aUude to the statement that in their young form they 

 closely resemble stellate spicules. This I cannot substantiate ; 

 closely as I have searched for transitional forms between 

 globates and stellates, I have never yet been able to find any, 

 any more than between stellates and bacilli. In thin slices 

 mounted in balsam I have been able to trace the globates 

 down to a young form, measuring something under y^Vu i^^ch 

 in diameter ; but even in tliis earliest stage it consists of a 

 vast number of minute trichites united into a central globule at 

 their inner ends. Its outline is spherical, owing totlie trichites 

 ending at the same distance from the centre; and it is enclosed 

 in a granular cell with a large young nucleus on one side, 

 which takes a deep tint with carmine. It thus differs from a 

 young stellate in just the same way as the stellate from a 

 bacillus, ^. e. by a great difference in the number of its rays. 

 As the globate increases in size, each trichite becomes longer, 

 thicker, and roughened over its free end, so as to resemble 

 closely an adult bacillus. The globate, indeed, miglit now 

 be well compared to a collection of bacilli, radiately grouped 

 and fused together at their inner ends. 



Classijication. — The generic distinction of Pachymatisma is 

 well founded, and is further supported by the character of its 

 oscular openings. 



In Geodia BarrettZy which we regard as an illustrative species 

 of the genus Geodia^ we likewise have an oscular tube ; but it 

 differs in a most important manner from that of Pacliymatuvia ; 

 for while the latter is separated by a sphincter at its base from 

 a common chamber below, in which the excurrent canals open 

 freely by unconstricted apertures,- the former, on the other 

 hand, is without the common chamber and the common sphin- 

 cter, and the excurrent tubes are severally and separately 

 sphinctrated as they open directly into the oscular tube itself. 

 In Geodia the oscular tube appears to result from the union of 

 a number of excurrent chones, like those of Isojjs'j in Fachy- 

 matisma it is produced by the over-development of a single 

 one. In Cydoniuni there are no oscular tubes, and both ex- 

 current and incurrent chones (if the distinction can here be 

 maintained) are covered with a cribriform or poriferous roof, 

 the very reverse of what holds in Isops^ where neither excur- 

 rent nor incurrent chones are so provided. Translating the 

 foregoing distinctions into a different nomenclature, it would 

 seem that in Isops we have a compound stock consisting 

 of a number of separate individuals, somewhat resembling 

 an ^5/r«(t-stock amongst corals ; in Geodia groups of these 



