326 DIVISION II.—COURSE OF DEVELOPMENT OF FUNGI. 
for ourselves of the only possible form of the parts when enclosed in the peridium. 
The cavity of the stipe is in this case also opened wide between the lobes; the 
somewhat more persistent outer layer of the gleba with its thin walls makes no 
difference in this respect. To arrive at the form of Phallus caninus or Simblum, 
we must have the stipe closed at the apex and projecting in a conical shape into 
the gleba, and the lobes must entirely disappear. We are not at present acquainted 
with any intermediate forms. But if none are found, we have sufficient material to 
show the family connection. Other genera than those already named are not taken 
into the comparison, because they are not clearly known or because it is unnecessary 
to adduce them. The reader is referred for them to the works which describe this 
group. ‘ 
Phallus and its nearest allies are the members of the series of the Phalloideae which 
depart farthest from the rest of the Gastromycetes. Clathrus is a link of connection 
between them and the Gastromycetes through the Lycoperdaceae. To see this we 
have only to compare its compound sporophores when ripening but still closed with 
those of Geaster, and the collenchyma-layer of most species of this genus, at any 
rate that of G. hygrometricus, with the receptaculum of Clathrus, The relation is 


FiG. 157. Colus hirudtnosus. FIG. 158. Aseroe rubra, Sketch of a mature recep- FIG. 159. Aseroßhallus. 
Sketch of a mature isolated recep- taculum which has issued from the peridium attached to it A ripe compound sporophore. 
taculum which has issued from the below, and bears the gleba in the middle of the radiati After M and Le- 
peridium. The black gleba hangs expansion. After Berkeley in Hooker's Journ. III, prieur, |. c., in note on p. 325. 
down under the sieve-like terminal Tab. V. Half the natural size. Natural size. 
plate. After Tulasne in Explor. 
Scientif. d’Algerie. Twice the natural 
size. 
brought out very clearly by the circumstance that there are individuals of C. 
eancellatus, in which the bars of the lattice-work, even in the ripe compound sporo- 
phore, are excessively broad and the interstices only narrow slits ; here therefore the 
receptaculum is a hollow body only slightly perforated. 
The connection between Clathrus and Geaster appears to me to be still more com- 
pletely established by the genus Mitremyces, which is chiefly American and still far 
from being thoroughly known. But I do not attempt to describe it here, for I have 
no sufficient account before me of the history of its development}. , 
6. A description of the genera Tulostoma, Poiysaccum, and Sphaerobolus may 
be given here by way of appendix, because they depart in a remarkable manner 
from the type of the groups which we have just been considering. But no special 
proof is required to show that they approach very near to them and especially to 
the Lycoperdaceae, and in this position may be regarded as the representatives of 
distinct and at present small divisions which have to be co-ordinated with the others. 

1 For the facts as at present known, see E. Fischer in Bot. Ztg. 1884. 
