hvilken den smelter sammen og danner et bredt Belte, 
Fig. 32, c, Fig. 34, a, der skjuler ganske et bagenfor lig- 
gende, transverselt Septum, til hvis Midte det er adhæ- 
reret, imedens der fra Beltets Siderande udgaa flere stærke 
Bindevævstraade, som fæste sig til Coenenchymets indre 
Vee, Fig. 32, d. Foruden at dette Belte er fæstet langs 
Midten af det transverselle Septum, er det tillige sammen- 
voxet til dettes nederste, fri Rand, Fig. 34, b. Fjernes 
dette Belte, saa træder Septum transversale tydeligt frem, 
Fie. 34, c; det er 2,5”” hoit, er fæstet paa begge Sider til 
Coenenchymets indre Veg, Fig. 34, d, og deler det nederste 
Rum i Stilken i to Dele, den ventrale og dorsale Kanal, 
idet de laterale Kanaler egentlig ophører der, hvor det 
transverselle Septum optræder, Fig. 34, e. 
Septum transversale er dannet af en dobbelt Binde- 
vævsmembran, der opad har en tyk Rand, Fig. 34,f, hvor- 
til det føromtalte Belte ogsaa er fæstet. Denne Dobbel- 
membran folder sig saa, at en Del slaar sig bagover paa 
Axeskeden, forlænger sig derefter nedover dennes forreste 
Veg, omgiver Skedens nederste Ende og danner ligesom 
en Sek, hvori denne ligger; den forreste Del af Membranen 
udgjør det egentlige transverselle Septum, hvis nederste, 
fri Rand er nogle Millimeter fra Stilkens Bund, saa at 
Enden af Axen sees nedenfor den, Fig. 34, g. Nedenfor 
Septum transversale er der selvfølgeligt kun et eneste Rum, 
hvori ikke alene den særdeles vide Dorsal- og Ventralkanal 
aabner sig men ogsaa Sidekanalerne, idet Bunden af disse 
sidste er gjennembrudt ved de mange Bindevævstraade, 
hvorved Septa fæste sig til Coenenchymvæggen. 
De dorsal-laterale Septa fæste sig indad paa Axe- 
skedens Sidevæg, Fig. 35, a, lige ned til Axens Ende, Fig. 
35, b; men naar de komme 1 Niveau med det Sted, hvor 
de ventral-laterale Septa afgive den omtalte Fold, der dan- 
ner Beltet, sende de en meget stærk Fold fra sig, Fig. 35, c, 
der bidrager væsentligt til Dannelsen af Septum transversale, 
Iie, BH, dk 
Unægteligt har det hayt sine Vanskeligheder at ud- 
rede disse temmelig indviklede Forholde i den bulbøse Del 
af Stilken, og det lader sig derfor let forklare, at Dr. 
Lindahl, der kun har havt et sparsomt Material, som han 
ikke engang tilstrækkeligt har kunnet benytte, har misfor- 
staaet disse Forholde. Han har ganske overseet baade det 
beskrevne Belte og Septum transversale; han udtaler sig 
nemlig saaledes: *Umbellula's skaft har 1 sin nedersta 
del en helt egendomlig bygnad, idet att åndblåsan utgöres 
af den utvidgade nederste ånden af ventralsidans hufvud- 
kanal, hvars båda sidoviåggar, de ventrala-laterala septa, böja 
sig in under de öfrige tre hufvudkanalerna, så att de bilda 
deras botten. Nåmnda septa göra dervid et starkt veck 
43 
inwards, in front of the sheath of the axis, and meets 
at the middle of the ventral surface, a corresponding fold 
from the other side, with which it becomes fused, forming 
a broad band (fig. 32, c, Pl. IX, fig. 34, a) that com- 
pletely conceals a transversal septum which les behind 
it, and to whose middle it is adherent; whilst, from the 
lateral margins of the band; several strong filaments of 
connective-tissue proceed, which are secured to the inner 
wall of the sarcosoma (fig. 32, d). Besides this band being 
adherent along the middle of the transverse septum, it 1s, 
also, conereted with the inferior free margin (fig. 34. 0). 
If this band is removed, the septum-transversale becomes 
distinetly prominent (fig. 34, ¢); it is 2.57” high, and is 
attached on both sides to the inner wall of the sarcosoma 
(fig. 34, d) dividing the inferior space m the stem, into 
two parts, viz. the ventral and the dorsal canals; because, 
the lateral canals really cease at the point where the 
transversal septum appears (fig. 34, e). 
The septume-transversale is formed, of a 
membrane of connective-tissue, with a thick superior mar- 
gin (fig. 34, f) to which the band previously spoken of, is 
also attached. This double membrane folds itself, in such 
manner, that a portion folds backwards on the sheath of 
the axis, then extends itself down along it’s anterior wall 
double 
and surrounds the inferior extremity of the sheath, forming, 
as it were, a bag, in which it is enclosed. The anterior 
part of the membrane forms the real transverse septum, 
whose inferior free margin is distant, a few millimetres 
from the base of the stem, so, that the extremity of the 
axis is observable below it (fig. 34, g). Below the septum 
transversale, there is, consequently, only a single space, in 
which, not only, the tolerably wide dorsal and ventral 
canals open, but, also, the lateral canals; as the bottom 
of these last, is perforated by the numerous connective- 
tissue filaments by which the septa are attached to the 
wall of the sarcosoma. 
The dorso-lateral septa are attached, inwards, to the 
lateral wall of the sheath of the axis (fig. 35, a), right down 
to the extremity of the axis (fig. 35, 0); but, when they 
come in line with the point where the ventro-lateral septa 
produce the fold spoken of as forming tbe band, they push 
forward a very strong fold (fig. 35 c) which contributes, 
principally, in the formation of the septum-transversale 
(fig. 35, d). 
The clearing up of these rather complex relations in 
the bulbous part of the stem, has, undeniably, had it’s diffi- 
culties; and, it is, therefore, easy to understand, how Dr. 
Lindahl, who had only slight material to work with, and 
which, he was not, even, in a position to sufficiently utilise, 
has misunderstood these relations. He has quite overlooked 
both, the band and the septum-transversale, which we have 
described. He states the following: “Umbellula’s shaft 
has, in it’s lower part, a quite peculiar structure, inas- 
much,that the terminal vesicle is formed by the dilated 
lower extremity of the main canal of the ventral side, 
through both of whose lateral walls, the ventro-lateral septa 
curve in, under the other three main canals, so, that they 
6 * 
