970 SUMMARY OF CURRENT RESEARCHES RELATING TO 



Morphology of Siphonophora.* — ^Prof, C. Cliun commences by dis- 

 cussing the structure of the pneumatophore, which is shown by recent 

 observations to be certainly a modified Medusa. He denies the accuracy of 

 Korotneff's statement that it contains any gastric cavity. The pneuma- 

 tophore consists, as is well known, of two lamellas, an outer one which 

 represents the continuation of the trunk, and an inner one which excretes 

 the air — this inner layer may be called the air-sac, and the outer one the 

 air-umbrella. Both consist of ectoderm and endoderm separated by a sup- 

 porting lamella. As the air-chamber is an invagination of the apical end 

 of the trunk, its inner surface is lined by ectoderm. In all the species 

 there is a constriction at the lower pole of the air-sac which may be called 

 the air-funnel. The lining ectodermal layer early forms a flattened epi- 

 thelium, and even in the embryo gives off a delicate chitinous lamella, 

 which forms a ring at the orifice of the sac ; this lining is, moreover, multi- 

 laminate ; the cells bounding the air-space are small and filled by a finely 

 granular protoplasm ; the underlying vacuolated cells gradually diminish 

 in size, and are so packed as to be polyhedral in form. 



There are a number of variations in the structure of the air-funnel 

 which appear to be characteristic of different species; the structure is 

 simplest in Apolemia uvaria ; the endodermal investment of the funnel and 

 the lowest part of the air-sac consists of long cells «radially arranged in 

 groups ; the ectodermal cells, which are separated from them by a delicate 

 supporting layer, form a thick multilaminate cushion. Part of this forms a 

 secondary investment over the chitinous ring, and it was this that Korotneff 

 mistook for a secondary stomach. In Stephanomia picta ( = Halistemma 

 tergestinum Glaus) the pneumatophore is provided with internal septa, and 

 these are swollen at their base owing to the entrance between the endo- 

 dermal of large ectodermal cells, which form a solid mass. 



In Physalia hydrostatica, the structure of whose pneumatophore has 

 never yet been completely understood, the number of septa varies, but is 

 ordinarily seven ; the so-called septal canals represent branched solid 

 " cellular tubes " which are formed of ectodermal cells, and which make 

 their way from the air-funnel between both the septa and the finely granular 

 ectodermal cells which grow into the lower fourth of the air-funnel. They 

 are the homologues of similar cells in Stephanomia. 



The remarkable structure of the pneumatophore of BMzopJiysa fiUformis 

 is due to the loss of the septa, while the ectodermal cellular cords between 

 them persist. 



With regard to the physiology of the several parts of the pneumato- 

 phore, it is clear that it is the function of the ectodermal lining and of the 

 secondary ectoderm to secrete air ; the latter is larger in proportion to the 

 size of the organ — in Physalia, for example, it forms a disc as broad as the 

 hand, though, strangely enough, it has never yet been noted by any observer. 

 The taking in of air from without is only possible in the Velellidse and 

 Porpitidse, where there are a number of air-pores ; they have no secondary 

 ectoderm or air-funnel, and their camerate pneumatopore is completely 

 invested by a thick chitinous layer. Though Phizophysa and Physalia 

 have an air-pore, it serves merely for the egress, and not for the entrance 

 of air. 



In investigating the morphology of this organ it is necessary to inquire 

 whether the pneumatojjhore is a characteristic of the higher Siphonophora, 

 or whether it has its homologue in a medusoid appendage of the Calyco- 

 phoridse ? Chun has shown that the definite nectocalyces of the latter are 



* Zool. Anzeig., x. (1887) pp. 511-5, 529-33. 



