ASTROSCLERA WILLEYANA, THE TYPE OF A NEW FAMILY OF SPONGES. 463 



then placed in an open dish in a thin solution of gum copal in chloroform. After 

 some days, when the solution had become thick, the specimens were put on slides, 

 with the thickened gum about them, and placed on the shelf of the warm water bath 

 (60° C.) to harden. When the gum was hard, slices of them were cut with the 

 fret-saw, one half being cut longitudinally, the other transversely. Each furnished 

 three slices. 



One surface of the slice was ground down smooth on a hone, and it was then 

 cemented to a slide with Canada balsam, with this surface downwards. When the 

 balsam was hard, the other surface of the slice was ground down until the section 

 was so thin that the edge began to break. It was then dried and covered with 

 Canada balsam and a coverslip. 



This method is that of von Koch, as given in Bolles Lee's Vade-mecum. It 

 may be worth while to mention that in grinding the sections, lumps of ice were kept 

 on the hone in order that the copal (which is soft at GO" C.) might be a.s hard as 

 possible. 



The structure of the hard and soft parts is well displayed in these preparations. 



The fourth Lifu specimen I have kept intact. 



Slices both vertical and tangential to the most recently formed part of the surface 

 of the Funafuti specimen, have also been treated by von Koch's method, but as the 

 soft tissues appear to be absent here, they are of value only as showing the structure 

 of the hard parts. The microscopic structure of the skeleton precisely resembles that 

 of the Lifu specimens described below. 



Skeleton. Microscopic characters. The fully formed skeleton is built up of a solid 

 mass of polyhedral elements whose surfaces are united together to the complete ex- 

 clusion of the soft parts. The elements vary considerably in size, 40/it being a frequent 

 diameter. An exceptionally large one measures 150/i in its larger diameter. Minute 

 rounded granules are often present at the centre, but the remainder of the element 

 consists of radially disposed crystalline fibres which terminate peripherally in contact 

 with the fibres of adjacent elements (Figs. 3, 4, 8, and 12). 



Mineralogical properties. My friend Mr A. Hutchinson, Fellow of Pembroke College, 

 has been so good as to examine a section and also some fragments of the skeleton 

 from the mineralogical point of view, and furnishes me with the following reports on it. 



"Under the microscope the section exhibits well-marked .spherulitic structure and, 

 when examined between crossed Nicols in parallel light, each spherule is seen to be 

 occupied by a black cross and a system of concentric coloured rings, forming an optic 

 picture similar to that observed when a plate of a uniaxal crystal, cut perpendicularly 

 to the optic axis, is viewed between crossed Nicols in convergent light. 



" Examined with a ;J-undulation plate of mica, these optic pictures behave as if 

 they were produced by negative uniaxal crystals. 



" The double refraction of the substance is strong. 



" The specific gravity, determined by suspending a small fragment of the substance in 

 bromoform, was found to approximate to that of Aragonite. Qualitative chemical analysis 

 showed that the substance is calcium carbonate. Magnesium was tested for with care, 



G2— 2 



