TAPEWORM PARASITIC IN THE STICKLEBACK. 125 
another work (28) he announced definitely that the larval forms 
of 7’. longicollis and 7’. ocellata occur encysted, with non-mature 
unsegmented bodies, in the livers of salmon and perch. 
Zschokke (26) found the unsegmented larva of 7’. longicollis in 
Salmo umbla. 
Similarly, von Linstow (14) states. ‘‘ Die Larva findet sich 
encystiert in der Leber desselben Fische, welche die erwachsene 
Tanie in ihrem Darm beherbergen.” 
Liihe (17) also inclines to this view. ‘ Als zugehérige Larve 
ist eine in der Leber von Perciden encystiert gefundene Cestoden- 
Larva betrachtet werden (¢).” 
“ Leuckart (13) reports a plerocercus found by Gruber in 
Cyclops serratulus which he believed to be the larva of Proteo- 
cephalus torulosa Batsch. From a comparison of his figure, 
which is drawn to scale, with measurements of P. torulosa, it 
seems that there is some foundation for this view” (La Rue (12)). 
To this larval host Lithe (17) adds C. strenwus. 
It may be seen from the above summary that no accurate 
experiments have been made with this genus. Most of the 
investigators incline to the view that the larve occur in 
the same host as the adult form, basing their opinion upon 
comparisons of the size and general shape of both forms. The 
connection 1s thus very loose. The infection would have to be 
by means of eggs. As there is no swimming-mantle, these do 
not float, but sink to the bottom; and since the sticklebacks in 
Edgbaston Reservoir feed chiefly upon Entomostraca, the com- 
parative rarity with which they would swallow eges could not 
account for the heavy infection observed. Moreover, the oncho- 
sphere would have to bore its way through the walls of the 
alimentary canal to the liver, encyst there, and then return to 
the alimentary canal again; for there could be no other way for 
it to infect both the liver and the intestine of the same host. 
This hypothesis 1s not very plausible and is not very well 
supported. At different times I have examined the livers of 
sticklebacks but without finding any trace of cysts, although 
nearly all the stages have been found in the alimentary canal. 
My own researches point to the conclusion that an inter- 
mediate host, Cyclops varius Lillj., 1s necessary for the full 
development of the egg, and it is through the final host 
swallowing the intermediate host, that the parasite obtains 
entrance to the alimentary canal of the former. This view is 
supported by the discovery of the plerocercoid of P. torulosa, 
previously mentioned*, 
Infection Haperiments. 
In studying the life-history of this parasite, I first began with 
the eggs. If the Cestode be adult or nearly so, directly it is 
removed from the intestine of its host, it begins to discharge the 
egys In a continuous stream from a slit in the ventral body-wall. 
* See Note ou pp. 1387-138. 
