THE MYXOSPORtDIA, OR PSOROSPERMS OF FISHES. 91 



Finally that, in at least one species, dispersal could hardly take place 

 by the myxosporidium is shown by Biitschli's observation ^ that in 

 Myxklium lieherl-iihnii that structure dies rapidly when removed from 

 its natural habitat (the urine of the pike) to even "indifferent fluids." 



(3) The myxosporidium, on the other hand, is the post-embryonic, 

 comparatively stationary, (jrowth-reproduction stage: There is little 

 reason to suppose that there is ever any migration from one host to 

 another during this stage. The evidence all points to the conclusion 

 that after (and probably soon after) its attachment to the host, the 

 valves of the spore separate, freeing the sporoplasm, which thencefor- 

 ward is known as the myxosporidium. Thus Lieberkiihn, Balbiani, 

 Pfeiffer, and Perugia have all seen the sporoplasm leave the si)ore 

 and exhibit amceboid movements. 



Now, if this view as to their relative functions in the life-cycle be 

 correct, the capsular filaments may conceivably serve in several ways. 

 First, they may serve as a flagelliform swimming apparatus, a view 

 that I think quite improbable, dispersal being more probably effected 

 by currents, etc. Second, they may (and this is probably their most 

 important function) serve for attachment.^ 



Further, if it be conceded that, after attachment, motion is necessary 

 to the spore, the filaments might easily subserve such function either 

 by a maximum extrusion, fixation of the tip, and a subsequent coiliug- 

 retraction (similar to that of the Vorticclla stem), the spore in this case 

 progressing " anterior " end foremost, or by a miuinuim extrusion fol- 

 lowed by fixation of the tip and progressive uucoiling-protrusion, the 

 spore in this case being pushed " posterior" end foremost. In Gliigea 

 anomaJa, which has but one filament, 50// long, motion could hardly be 

 effected in the latter way. But such motion is easily conceivable with 

 the 2-capsuied {MyxoboJus, etc,.) spores; and if it were admissible to 

 suppose that the final lodgment iireliminary to reproduction is ever 

 effected by the spore and not by the myxosporidium, the latter being 

 liberated and growing in situ (a view which, however, the present 

 evidence tends to negative), this backward motion would be the best 

 ])ossible for inserting the .spore under a scale, especially for those species 

 provided with a tail, wiiicli latter structure would form an efficient 

 guide to such insertion. I incline, however, to the view that the func- 

 tion of the filament is attachment, and that the motion necessary for 

 the attainment of a place for reproduction-encystment is effected by 

 the liberated myxosporidium. 



1 Ztschr. f. wiss. Zool., 1881, xxxv, p. 639. 



2 Perfectly consonant with this view is tlie observation of IJiitsclili (Ztaclir. f. wiss. 

 Zool., 1881, xxxv, p. 635) that the lilanients are extruded in spores which are pre- 

 served a long time in water. For we thus see the floating spores ready for instant 

 attachment to any object with which they may come into contact. A possible cause 

 for such extrusion might perhaps be found in osmotic pressure (pre])on(lerant endos- 

 juosis from the surrounding water) from within. At any rate, it is diflicult for me to 

 attribute the rupture of the shelled-out cyst observed by M. Thelohau (see p. 221) 

 to any other cause. 



