37 8 DISCOVERY REPORTS 



of integration (which has still to be fully analysed) is in evidence. It appears that Forskalia can swim 

 forwards or backwards by altering the shape of the mouths of the swimming bells ; the shape of the 

 bell mouth in Forskalia appears to depend on the degree of contraction in two groups of radial muscle 

 fibres in the velum, an arrangement unique to this group. Phases of activity or inactivity are common 

 to the whole assemblage of swimming bells, but when active the bells beat at different frequencies 

 after the first one or two beats. Certain observations point to the possibility of inhibition as a co- 

 ordinating mechanism in the activities of the swimming-bells (Mackie, cited by Hardy, 1956). 

 The observations in question were, however, made on a few, rather imperfect specimens, at a time 

 when the author was not fully aware of the issues involved, and it would be of great interest to repeat 

 the work. 



In the case of certain Calycophora (for example, Hippopodius, Chelophyes), histological study 

 provides clear evidence that in contrast to the arrangement in Forskalia, where two median exum- 

 brellar nerve tracts connect the marginal rings of the bell with the stem, there is in these genera no 

 direct nervous connection between the stem and bells. What is remarkable, however, is that in spite 

 of the histological absence of nerve connectives, observations on living calycophores suggest that some 

 sort of integration does, nevertheless, exist. 



The problems of nervous co-ordination in the siphonophores are therefore considerably more 

 complex than study of a simple form like Physalia would tend to suggest. 



HISTOLOGY 



1. Introduction: The Siphonophora compared with other Hydrozoa 



The main features in the histology of the siphonophores were established by the end of the nineteenth 

 century. Good comparative accounts were given by Chun (1897, 1902) and Schneider (1902), the 

 two workers who contributed most to the actual investigations. There has been little subsequent work 

 of importance. 



In their basic microscopic structure, the siphonophores bear a close similarity to other Hydrozoa. 

 However, the tissues are often brought to a higher degree of elaboration and histological perfection 

 than elsewhere in the Class. This is true of the striated muscle in the swimming-bells of certain 

 Calycophora, where Q, J and H bands are clearly distinguishable ; of the nervous system in the bells of 

 Physonectae, where exumbrellar tracts are developed ; and of the mesogloea in genera such as Hippo- 

 podius, Chelophyes and Diphyes, where the exquisite moulding of the swimming bells is achieved by 

 means of an elaborately orientated system of mesogloeal fibrils. One must also mention the compli- 

 cated mechanism of the nematocyst batteries in such forms as Fray a sp.n.* (Korotneff, 1884), which 

 is without parallel in the entire Phylum. 



In addition to the tissues common to all hydrozoans, the siphonophores (except for the Calycophora) 

 possess gas-secreting tissue, for which the only known counterpart in the Class is the basal disk in 

 Hydra (Kepner and Miller, 1928; Kepner and Thomas, 1928). The Calycophora show two tissue 

 regions for which no counterpart is known : the capillary network in the radial canal system of the 

 swimming bells, and the sack-like ' somatocyst ', whose cells show a great capacity for changes in 

 volume. The functions of these regions are unknown. 



There is one other striking feature in the histology of the siphonophores which should be men- 

 tioned. Although the two marginal nerve-rings are present, the swimming bells lack a subumbrellar 

 nerve-plexus. Conduction across the muscle-sheet must therefore be myoid. This and certain other 

 evidence suggest the possibility that, in hydrozoan medusae generally, the swimming beat, though 



* Identified by Totton as Stephanophyes superba Chun. 



