384 Transactions of the Society. 



Hardly any two specimens exhibit an identical shape or 

 external appearance. This diversity is due to the methods of 

 construction and the material employed. The apparent mode of 

 construction is to select a number of long slender spicules often 

 2 or 3 mm. or more in length. These are placed like tent poles 

 at various angles about 0*5 mm. apart, forming a rough open-work 

 figure enclosing a central space between the points of intersection 

 of the poles. The open spaces in the wall are then filled in with 

 shorter fragments of spicules carefully selected for length, so as 

 just to fill the required space. The animal thus secures the nearest 

 possible approach to a spherical chamber obtainable with the 

 material employed, the salient angles being the points where two 

 or more of the " tent-poles " join. The long spicules employed as 

 " tent-poles " project irregularly all over the surface of the test in 

 perfect specimens, and probably serve a secondary purpose as 

 catamaran spars in supporting the animal in the surface layer of 

 ooze. They are, however, very fragile, and are frequently more 

 or less damaged, if not destroyed, in the process of cleaning the 

 dredged material. 



The internal cavity of the test is quite devoid of projecting 

 spicules and is not coated with cement. 



As a rule spicules only are employed in the construction of the 

 test, but occasionally a minute grain of sand or flake of mica is used 

 to close the little corner space where two or more " tent-pole " 

 spicules meet at an acute angle. Still more rarely this angular 

 space is left unclosed, thus constituting an aperture to the test. 

 The presence of such apertures must, however, be regarded as 

 abnormal. As in Psammosplixra fusca (Schulze) there is not any 



EXPLANATION OF PLATE VI. 



Fig. 1. — Technitella legiivicn Norman. Showing the differential arrangement of 

 the sponge-spicules in the external and internal layers. The 

 interstitial cement is represented by the shading. X 140. 



,, 2. — Fsammosphxra rustica sp. n. x 40. 



,, 3. Ditto. An abnormal triple specimen, x 40. 



,, 4. Ditto. Detail showing the method of construction in a " panel " of 

 the test. The central space has been filled in by a fragment 

 of a triaxial sponge-spicule. The interstitial cement is re- 

 presented by dark shading, x 140. 



„ 5. — Psammosphiera Boivmanni sp. n. x 40. 



,, 6. — Marsipella spiralis sp. n. x 40. 



,, 7. Ditto. Detail showing the method of construction of test, x 140. 

 „ 8. — Marsijwlla ci/lindrica Brady, x 40. 



„ 9. Ditto. Detail showing the loosely-constructed terminal crown. The 

 constituent spicules are not embedded in cement, x 95. 



Note. — AU. the figures in this plate are drawn from transparent balsam- 

 mounted specimens, and for the purpose of clearness the 

 focal plane has been more or less disregarded in figs. 1, 4,7, 

 and 9. 



