Dr. IT. RaufF on the Genus Hindia, Dune. 173 



to the same canal; the meshes in every two laterally contiguous 

 rows always stand alternately, by which a very elegant image 

 is produced. In certain good parts of the preparation it is 

 clear that the nodes (central points) of the laterally contigu- 

 ous elements are also placed alternately, and that the union 

 is effected in this way : the widened and denticulated head 

 of one arm meets the convex and likewise denticulated main 

 portion of the arm, lying in the same face of the canal, of that 

 laterally contiguous element whose nodal point does not lie 

 in the same transverse section with the first one, but about 

 half a nodal distance (half the height of a mesh) above or 

 below it. Thus the arms 1 and 2 of one element (fig. 1 

 and tube I. in fig. 2) always unite respectively with 2 and 1 

 of the neighbouring elements on each side. At the same 

 time the arms of all the elements lying on the same radius are 

 parallel in direction, so that two arms, 1 and 2, not belonging 

 to the same element but to laterally neighbouring ones, always 

 come to lie in the same face of the canal-wall. Moreover, 

 all the convex sides of the arms are always turned in the 

 same direction, namely outwards towards the surface, and all 

 the concave ones towards the central point. For the canals 

 represented in fig. 2, therefore, the central point of the sponge 

 is to be sought in their prolongation upwards. 



As the heads do not meet, but always the head of one arm 

 with the main portion of another, the trabeculaj of the skeleton 

 (measured sideways between the meshes) when the sutures do 

 not show distinctly or are obliterated, or under only a low 

 power, appear much thicker than they are in reality, in fact 

 often about twice as strong. 



If we now suppose that, as everything seems to show, the 

 third arm is equivalent to the first two in position, in structure, 

 and generally in its relations within its own element and to 

 the neighbouring ones, and that the fourth arm is rudimentary 

 or deficient (its development, however, would by no means 

 alter the principle of structure here put forward, the assump- 

 tion of such a development being only unobserved, whilst 

 in point of fact, as indicated in fig. 1, the presence of a free 

 unattached stump (4) projecting into the mesh from the 

 nodal point could frequently be recognized),-^if we make 

 this supposition with regard to the third arm, the skeleton, 

 by the constant uniformity of the mode of union of all its 

 members, will naturally form itself into nothing but six- 

 sided radial tubes, each contiguous pair of which will always 

 have a common wall. 



In fig. 2 the three diagrammatically represented hexagonal 



