which we suppose to belong to the same giant-cells as those in the air-funnel. In one of its 

 folds we find the chitinous layer {f.p}'.) enveloped by the secondary ectoderm. 



Further on (on glass 1 1 III 3 sketch fig. 127 III) we find that the air-funnel has preserved 

 its branched appearance and comes into closer connection with the opposite side, the wall of 

 the pneumatosaccus. This happens by the union of the chitinous layer of the former with the 

 chitinous layer of the latter. The entodermal layers have united too and continue in one 

 unbroken line. The air-funnel contains the same chitinous substance imbedded in its lower part. 

 This substance is more voluminous and becomes also dichotomously branched. The secondary 

 ectoderm with the chitinous substance in its folds, the remnants of giant-cells both in the 

 secondary ectoderm of air-funnel and pneumatosaccus have remained the same as in sketch 

 125 II. There is no septum connecting pneumatocodon and pneumatocyst. 



In sketch IV (PL XVTI, fig. 128 glass 11 111 7) the fulcrum breaks through and we find 

 now a definite continuation of the wall of the pneumatosaccus and the pneumatochone. This 

 aperture appears suddenly, the chitinous layer gets thinner and from one section to the next 

 (glass II III 6 — II III 7) the breaking takes place. 



The other layers follow the ectoderm of the air-funnel meeting the same layer of the 

 pneumatosaccus; the same is the case with the entoderm and the fulcral layer. Also in this 

 sketch the ectoderm in the lower part is supposed to consist of many layers. Lying immediately 

 aeainst it we find that the chitinous substance in the air-funnel is distinctlv branched, ends on 

 both sides in two club-shaped enlargements {cp}'-) which are connected by a thin thread-like 

 stripe of chitinous substance. This stripe or ribbon faces now the chitinous ribbon of the 

 pneumatosaccus. 



Between those two club-shaped enlargements and also on the proximal side fusing with 

 the ectoderm, which is situated at the bend of the two side-branches, we find a tissue which we 

 cannot identify entirely with the secondary ectoderm. It is of course very much altered by 

 preservation, but seems to us to contain fibrous substance intermingled with secondary ectoderm 

 cells. They constitute, therefore, probably the matrix-cells of these thick, hard, chitinous club-shaped 

 enlargements. We have on purpose given them the same colour as the secondary ectoderm, 

 though somewhat darker, but the boundaries of these cells are not clear in any of our sections; the 

 demarcation between the light and the darker green is therefore to be considered an absolutely 

 hypothetical one. In sketch IV (PI. X\'II, fig. 128) there are no further pecularities, except for 

 the occurrence of a new septum between pneumatocodon and pneumatocyst. 



Sketch V (PI. XVII, fig. 129 V) was taken from a section immediately following the one 

 from which we tried to reconstruct the whole in the diagrammatic fig. 124 of Plate XVII. It 

 needs therefore little explanation. A further connection of the air-funnel and the pneumatosaccus 

 has now arisen by the union of the inner chitinous layer of the latter with the club-shaped 

 enlargements especially with their connecting band. 



We have now come to the stage which makes the resemblance of this pneumatophore 

 a striking one to those described and figured by Chun 97a. A comparison of the sketch V of 

 Plate XVII, fig. 129 and fig. 124, Plate X\TI with Chun's figures given of the pneumatophores 

 of PJiysophora and Athorybia shows how easy it has now become to reconstruct the structure 



SIBOGA-EXI'EDITIE IX. 



