210 DIVISION II.—TOURSE OF DEVELOPMENT OF FUNGI, 
hyphae begin to shoot out from the sterile sister-branches of the archicarp and from 
the whole of the rest of the basal region of the rosette to form the envelope-portion 
of the sporocarp (Fig. 98), consisting of a receptaculum enclosing the group of 
archicarps and antheridia and the hypothecium with the paraphyses, which latter 
always rests free on the receptaculum without an enveloping wall. The dis- 
tribution of the ascogenous hyphae and asci and the gradual multiplication of the 
latter between these elements of the envelope are essentially the same as in Ascobolus. 
The paraphyses form at first a conical tuft on the hypothecium, which generally 
broadens out into a disk through the introduction of new elements. The receptaculum 
becomes a comparatively large thick large-celled pseudo-parenchymatous disk covered 
with rhizoids, and between its elements those of the primary rosette are inclosed, and 
are at length indistinguishable (Fig. 99). The antheridia continue longest visible and 
indeed almost unaltered, being very full of protoplasm, and take no part in the formation 
of the envelope. 
7. In Sordaria among the Pyrenomycetes especially S. (Hypocopra) fimicola and 
in Melanospora parasitica the course of development in the perithecium, according 
to Gilkinet’s and Kihlman’s researches, is essentially the same as in Ascobolus, of 
course with certain specific differences, and with differences of conformation corre- 
sponding to the difference between Discomycetes and Pyrenomycetes. The archicarp 
is a spirally coiled cell-row, though in Melanospora it is sometimes almost straight. 
Antheridial branches are less plainly seen in Melanospora, or at least are not sharply 
distinguished from the incipient filaments of the envelope, which here too grow 
close round the archicarp soon after its formation. In Sordaria the growing 
archicarp divides by transverse walls into numerous cells, and ascogenous hyphae 
sprout from the great majority of these cells; whether any portion of the archicarp 
takes no part in their production could not be determined owing to the early gelatinous 
disorganisation of the walls of all its cells. In Melanospora only one or two cells in 
the middle of the large archicarp become ascogenous, the rest being disorganised 
and afterwards partly ejected in this state from the orifice of the young perithecium. 
These one or two ascogenous cells develope by successive bipartitions in varying 
directions into a parenchymatous body, the many cells of which are full of protoplasm 
and subsequently produce the asci. These are arranged in Sordaria nearly parallel 
to one another in a thick tuft, in Melanospora they form a nearly spherical body, and 
their apices converge towards its middle. In both cases the many-layered pseudo- 
parenchymatous wall of the perithecium is formed from the weft of hyphae of the 
envelope. It is a spherical body at first closely surrounding the future group of asci 
on every side, the neck and the canal of the ostiole being formed in it later. There 
are no paraphyses standing between the asci; these are placed separate and beside 
each other on a surface of insertion which covers a part of the interior of the sphere. 
On the side left free from the paraphyses there is a narrow empty space between the 
wall and the ascus-group, but this is soon filled with a large number of closely packed 
hyphal branches, which grow into it as perzphyses from the wall, converging radially 
till they touch one another. The group of these which is mostly turned away from 
the asci then grows vertically towards the wall in the direction of the neck which is 
to be, and so on to the outside through a hole in the wall, and thus forms the 
commencement of the neck, which may lengthen out considerably, and in Sordaria is 
