MICROSCOPICAL STUDIES. 37 



coriacea, budding, and branching, and anastomosing long continued, 

 produces a colonial network having a long ramifying paragaster 

 with numerous oscula. An intermediate form, Ascaltis botryoides, 

 helps us better to understand the change, for here the buds while 

 remaining attached to the parent and with their paragasters in free 

 communication with the parental one, have each a separate osculum. 

 Thus in an old colony, we can trace distinctly of how many indi- 

 viduals it is composed, by counting the number of branches, for 

 each possesses a single osculum, the sign of the sponge unit. 



A foreign species furnishes another interesting complication 

 shown in Fig. 2. The walls become pushed out into numerous 

 radial tubes, into which the flagellated coating of the paragaster is 

 extended, and into which the pores open. Thus the extent of flagel- 

 lated surface and of the pore area is largely increased. The name 

 Chambered Ascon is applied to this form. 



The next sponge type given us by Hseckel, the Sycon, is a 

 natural outcome of the Chambered Ascon, and in simplest organization 

 has the same structure minus the flagellated lining of the primary 

 paragaster. The cells of this become practically identical in form with 

 those of the ectoderm by loss of flagella and collars. Thus the Sycon 

 may be defined as a Chambered Ascon where the flagellated cells are 

 restricted to the radial tubes. Figs. 3 and 4 graphically exhibit 

 this change. These two types are practically restricted to the 

 division of sponges possessing a skeleton of calcareous spicules, which 

 is thus marked out as the lowest or most primitive division of the 

 Porifera, and consequently also of the Metazoa. 



. The third and highest type of sponge canal system, called 

 by Hseckel the Leucon from its being characteristic of the family 

 Leuconidse, and by Sollas the Rhagon, is accompanied, or rather 

 caused, by a great development of the middle layer of the body, the 

 mesoderm. This occasionally reaches immense proportions (see Fig. 8) 

 and as the paragaster does not share in this increase, the rather 

 dwindling, it is obvious that the pores must have, superadded, long- 

 canals to enable the water to pass through the thickened walls into 

 the paragester, or even into its out-growing chambers. One origin 

 of the Rhagon is probably from the Sycon, as shown by the hypo- 

 thetical Figures 5 and 6 ; Fig. 5 is practically a thick-walled Sycon, 

 the paragaster pushing into the mesoderm fairly large rounded 

 chambers, henceforth to be called ciliated chambers. Narrow tubes, 

 incurrent canals, connect these with the pores on the surface, and 

 either one or several may serve each chamber. The flagellated form 

 of endoderm is entirely confined to the chambers. The next step 

 is where the mesoderm still thickening, the chambers lose their 

 direct connection with the central cavity and become connected by 



