300 
DORIS L. MACKINNON. 
forked and otherwise 'Reformed. Four flagella^ 
arise from two basal granules at the extreme anterior 
end in two groups of two; they are sub-equal in 
length/ and are considerably longer than the body; 
there is no Schlepp-geiss el.^^ A cytostome lies 
between the basal granules. Periplast relatively 
thick and rigid, raised into numerous folds or ribs, 
which run in a direction approximately parallel to 
the long axis. Cytoplas m finely al veolar, contain- 
ing numerous ingested bacteria. Axostyle usually 
present, but feebly developed. Nucleus immediately 
behind the basal granules; sometimes spherical, 
more often pear-shaped, with large karyosome 
surrounded by a clear zone, which is surmounted 
by a group of chromatin granules. Extra-nuclear 
granules of chromatin (?) may also occur. 
Dimensions, 7 ^ to 1 5 /x x 4 /x to 6*5 ju. 
Whether the Polymastix found in Tipula is to be 
considered as a species distinct from the type-species, 
Polymastix melolonthae (Grassi), it is difficult to say. 
Hamburger considers that the same species lives both in 
Melolontha and in Cetonia. I have recently had the 
opportunity, through the courtesy of Dr. Carlos Fran 9 a, of 
examining a preparation of Polynia stix from the gut of a 
larva of Oryctes nasicornis, and it appeared to me 
indistinguishable from the flagellate of Tipula. The 
evidence, then, is in favour of the view that a number of 
insects of similar feeding habits may be parasitised by 
Polymastix melolonthae (Grassi). 
The following observations have reference to the parasite 
as found in Tipula. 
^ It is to be noted that certain undoubted Trichonymphids, such as 
Devescovina striata, have only four flagella. 
^ Authors — Hamburger, JoUos, Alexeieff — state that the flagella are 
unequal in length. I think that this is so only in individuals that have 
recently undergone division ; in such a case two flagella may be re- 
grown, and for some time these will appear shorter than the rest (vide 
infra). 
