45 



2. Description of the anatomy of a typical example of tr o pus tubulifera. 



For this purpose I take specimen 1365- A longitudinal section through the axis of a 

 process shows us a large central canal opening with a wide mouth at the top (PI. V, fig. 1 1 ). 

 In the canal we see with the naked eye numerous rugae; between these we observe some 

 small openings which partly communicate with longitudinal canals in the wall of the tube, as 

 indicated in the drawing by the sonde. Thin longitudinal sections, examined with a low power 

 teach us that these secundary longitudinal canals have very different diameters and that some 

 of them communicate with the main central canal even at the very top of the process (PI. V, 

 fig. 18). In addition to these canals the naked eye discovers numerous very small openings, 

 partly clispersed irregularly, partly in a series close under the sponge surface. The substance 

 of the parenchyma looks consequently rather compact ; something like a cortex is hardly visible. 

 Thin transverse sections show, mutatis mutandis, the same distribution of canals (PI. V, fig. 19). 

 If we look at the sponge surface with a pocket lense we observe innumerable small apertures 

 (PI. V, fig. 21). These are, at least partly not incurrent but excurrent apertures, procts. It is, 

 however, probably that they represent partly, stomata. Since we found in 1945 no difference in 

 size or structure between stomata and procts we cannot be astonished to find the same in the 

 specimen under description. The most striking thing in 1365 is that the large central canal 

 seems not to be an excurrent canal and consequently the large opening at the top of the 

 process not an "osculum". By and by I will give several arguments for this more or less 

 perplexing statement. One argument I will give at once, being based on the examination of a 

 series of sections. 



In the shape of the mastichorions we have an index to determine at which side the 

 water enters and at which it leaves these organs. Pekelharing and the present writer showed 1 ) 

 in which way the choanocytes may act as valves. According to our views the opening, which 

 is generally considered the efferent one and which is known as apopyle, must forcibly be the 

 excurrent aperture of the mastichorion indeed. Hence the canal which communicates herewith 

 is an excurrent one ; the others are consequently of an incurrent nature. Now I found that 

 canals, for the above reasons declared to be incurrent, communicate with the large central 

 cavity. We saw that in this cavity openings are found of various diameter. But in addition to 

 these — visible to the unassisted eye — other ones occur, some of which are very small and 

 are only seen with the aid of the microscope. All these openings lead into a system of canals 

 and lacunae which terminate in ultimate incurrent canals (PI. VII, fig. 1,2). Around the final 

 branches are situated mastichorions, which on the other hand communicate by wide apopylae 

 with excurrent canals. Again we find here some of these canals formintr an interdieitatincr 

 system. And again we find, as in 1945 the mastichorions situated in groups. In those cases 

 these groups occur near the central tube as illustrated in fig. 1 on PI. VII, the excurrent canals 

 unite into some much wider canals {exc), from which again narrower canals start, which ramify 

 and finally open with small, but numerous procts on the sponge surface. 



1) 1898 u p. 15. 



