96 



gradually into the outer wall of the pneumatophore {pd) and becomes the outer wall of all the 

 younger and mature appendages on the basal part of the corm. The chitinous layer of these 

 appendages is continuous with the chitinous layer of an interior cavit\- indicated by a green 

 colour. We find this to be the air-funnel {J>.c.) and as we find a great many granulations in the 

 interior of this cavity we suppose these belong to the giant-cells ; the latter are again formations 

 of the gas-secreting secondarj- ectoderm, and we think we are justified in supposing the cavity 

 of the air-funnel to be filled up entirely by secondary ectoderm. The giant-cells in this first and 

 in all other sketches are represented diagrammatically by small black dots. It is only in very 

 few cases that we were able to find any contour to these cells. Some of these have been figured 

 enlarged on PI. XVIII, fig. 140. 



The pericystic cavity is lined by an entodermal layer, which in the appendages has many 

 folds. A magnified sketch (PI. XA^Il, '[\<i:[^. 134) of one of these folds shows the extraordinary 

 resemblance there is to the drawings given by Haeckel of Rlwdalia (88b PI. V, fig. 27). 



On the side of the pneumatosaccus we see (PI. X\TI, figg. 125 — 133/. ,ra^f.) of course 

 fu'st the entoderm, then the very irregularly shaped thick chitinous layer which in Archangelopsis 

 seems to have attained extraordinary thickness. 



The ne.xt layer is a chitinous one again which in this first sketch does not lie quite 

 against the elongated strip of secondary ectoderm. It constitutes the pneumatocyst ; the folds of 

 this layer into the secondary ectoderm would perhaps suggest that also in this point Archan- 

 gelopsis is primitive. As is said in the note on p. 94 the secondary ectoderm originalh' in many 

 instances secretes a chitinous layer (see Woltereck 1905) and it seems we find this primitive 

 .stage again in the cushion-like secondary ectoderm of our specimens. Of course it may also be a 

 result of contraction of tissues but we found the same in our three specimens of Archangelopsis 

 (PI. XVII, fig. 126 II/.;^v., fig. 127 HI/./;'., fig. 128 IV /./J'.). 



In the other sketches (PI. XVII, figg. 126 — 133 II — IX) we have left out the young 

 appendages. We sketched only the air-funnel, showing here and there the transverse septa which 

 connect it with these appendages (II, I\\ VI, etc. s.a.a) and part of the pneumatocodon on 

 the rieht and left sides of this air-funnel. 



The second diagram (PI. XVII, fig. 126 II) thus shows the outer wall or pneumatocodon 

 {p.o.) on both sides of the now dichotomously branched air-funnel. A septum [s.a.h) connects 

 the pneumatocodon and the pneumatosaccus, and is, as is seen by its position, different from 

 the one incHcated in PI. X\'1I, h^^. 125 I. 



The air-funnel has not only become branched but has also increased in size, as is shown 

 !:>)■ the outer wall of the pneumatosaccus, coming closer to the entoderm of the opposite side. 

 The ectoderm of the branched part was in some parts well-preserved and proved to consist of 

 many layers in its basal part. In this part, quite near to the thick ectodermal layers we see the 

 first appearance of a chitinous substance; the ectoderm consisting of more than one layer has 

 probably contributed to the formation of this chitin. In the air-funnel we find again granulations, 

 the divided nuclei of giant-cells and .secondary ectoderm. The secondary ectoderm {sec. exy.) 

 facing the interior cavity or pneumatosaccus has increased in size as compared with sketch I, but 

 preserves its elongate cushion-like appearance. It contains also cavities filled with granulations, 



