AMPULLAE 355 



stages of regression from gastrozooid to gonopalpon (non-feeding gonozooid) can be seen when 

 tracing the development of a gonodendron by successive budding or replication. 



All palpons, it would seem, are reduced secondary gastrozooids (gonozooids), specialized for pur- 

 poses other than feeding. It is in Apolemia uvaria that they most nearly resemble gastrozooids, because 

 in this species the tentacle of the gastrozooid is unbranched and looks like a palpacle. The repeated 

 sac-like protuberances from the stem of Rhizophysa — the buds of the gonodendra — show little 

 similarity to palpons, though the side branches terminate in palpons; but the intricately branched 

 gonodendra of Physalia clearly originate from the peduncles of secondary gastrozooids. 



It has always been a puzzle to understand the appearance of a partial ring of palpons round the 

 base of the protozooid at an early post-larval stage in the development of physonects. But in the 

 light of what has now been learned about the branching system in Physalia, it has become easier to 

 homologize the process of budding of successively more reduced daughter polyps from the bases of the 

 gastrozooids in Physalia with the phenomenon to which I have just referred, namely the budding of 

 a partial ring of palpons. In Forskalia the whole gonodendron buds from a basal palpon; in Marrus it 

 probably does so as well; in Athorybia* there is a palmate arrangement of eight or nine palpons arising 

 from the pedicels of the gonodendron, but it would be necessary to study growth-stages to determine 

 the minute structure. 



Ampullae 



Haeckel (1888) — to whose account of Physalia most students hitherto have been likely to refer — used 

 the same term for true palpons and for the basal ampullae of the tentacles. For a long time I have held 

 the view that these ampullae, containing as they do in their walls innumerable nematoblasts, were 

 homologous with basigasters or nematoblast-nurseries — otherwise missing in Physalia, except in the 

 protozooid — which have become separated from the gastrozooids, just as the tentacles have. Schneider 

 (1898), I find, came to the same conclusion. But for a time I had to modify my interpretation because 

 on two occasions I found young growth-stages of ampullae with papillae inside, and a terminal part 

 that resembles a gastrozooid instead of the tip of an ampulla. My final conclusion is that an ampulla 

 is simply the hypertrophied basigaster of a gastrozooid. This is the sort of condition found normally 

 in the young growth-stages of gastrozooids of all siphonophores, where the basigaster is relatively 

 enormous, and the as yet undeveloped terminal part appears only as a slight projection from it. 



At the proximal end of a tentacle, particularly a young one, it can be seen that the last few nema- 

 tocyst-' heads ' budded contain no nematocysts. In this region, nematocysts can be seen in the 

 ectoderm of the muscular lamella where it connects the ampulla to the tentacle : these nematocysts 

 appear to be migrating to the 'heads'. There is in this region a sudden transition between 'heads' 

 devoid of nematocysts and ' heads ' well armed with them. As the ampulla and tentacle increase in 

 size the ' heads ' become bigger. Migration of nematocysts into the ' heads ' takes place only in the 

 restricted area already described. The original tip of a tentacle has, naturally, small ' heads '. 



A fact of great evolutionary interest is that the protozooid which has only a small tentacle, has no 

 ampulla, but instead a very inconspicuous basigaster of a generalized type. We have here then an 

 evolutionary phenomenon, found also in physonects like Agalma elegans, where the protozooid and its 

 basigaster remain at an earlier phylogenetic level, while the basigasters in the secondary gastrozooids 

 are more highly evolved. Instead of being able to deduce the evolutionary steps from a study of 

 fossils, we can do so, in bud-colonies of this sort, from a study of successively produced buds. 



* These palpons do not open directly as stated by Haeckel into the common stem cavity. It is necessary to anaesthetize the 

 delicate creatures to cause relaxation of the muscles before one can really observe their structure. Even so, the air in the float, 

 which always rises to the surface, makes it almost impossible to observe the under-surface until the air has been replaced by 

 liquid. Today of course we have what Haeckel had not — binocular dissecting microscopes. 



