1028 Butler.-—On Allomyces , <2 Aquatic Fungus . 
becomes a reproductive organ. The nearest approach to it is found in the 
Saprolegniaceae, in several of which the old mycelium may segment to 
form ‘ gemmae ’ and in one case, Saprolegnia torulosa , de Bary, segmenta¬ 
tion is normal. This segmentation is, however, of quite another order to 
that of the Leptomitaceae (cf. Maurizio, ’ 96 ). In the latter group the 
segmentation is not a phenomenon of budding and only rarely is the 
interruption between successive filaments complete; usually a central 
passage, through which the protoplasm of adjacent segments is in direct 
communication, is left. In Rhipidium continuum and some plants of 
Gonapodya polymorpha the segmentation is obscure or lost. In Apodya 
lactea, however, the passage may be entirely closed by a deposit of cellulin ; 
the process has been fully described by Pringsheim (’ 85 ). In Gonapodya 
something similar occurs, according to Thaxter, while Petersen’s figures 
(TO, pp. 533-4) show complete interruption. In the cases examined by 
Pringsheim the passage is blocked by one or more cellulin corpuscles which 
move into it from the adjacent segments and become fused with the walls. 
In the new Fungus the process is different, a distinct membrane being 
thrown across the cell by an ingrowth or deposit which begins at the 
periphery and gradually extends to the centre (Fig. Nevertheless, 
in its formation it is probable that the cellulin granules take part. These 
bodies are numerous and distinct and become massed at the point where 
the septum is developing (Fig. 17, a,h). They are ultimately divided into 
an upper and a lower group by the formation of the septum. The 
resemblance to what occurs in the differentiation of the sporangium in 
Saprolegnia, as described by Rothert (’ 90 ), is considerable, though in the latter 
case the whole septum appears simultaneously and not by an ingrowth 
from the periphery. It is clear that the case represents a more advanced 
stage in the segmentation of the Leptomitaceae and is not comparable 
to the true septation of the higher Fungi. The resulting septum is usually 
thin and similar to the cell-walls, but it is sometimes thick and irregular, as 
in Apodya (Fig. 1 J,e). 
The sympodial axes of the fertile hyphae, the basipetal development 
of the sporangia, and the position of the reproductive organs are minor, but 
interesting points of affinity between the Fungus above ‘described and the 
Leptomitaceae. Rhipidium and Apodachlya should be compared on these 
points. 
The peculiar resting spore is more important and its homology with 
that of Blastocladia evident. Apodachlya , an undoubted member of the 
Leptomitaceae, is the only other Oomycete with thick-walled resting spores 
which can at all be compared with these. The conidia of Pythium and 
the higher Phycomycetes are merely modified sporangia, and hence belong 
to another spore category. 
In view of the above characters, the inclusion of the new Fungus amongst 
