HISTOLOGY 379 



originating in the marginal nerve rings, may be conducted across the muscle-sheet independently of 

 the nervous system, the latter serving simply to negotiate radial responses between the margin and the 

 manubrium. The absence of the subumbrellar plexus in swimming-bells and in such forms as 

 Eucopella (Lendenfeld, 1883), may be correlated with the absence of a manubrium. A nerve-plexus 

 has never been found in the velum of a hydrozoan, and here too the muscle must presumably conduct 

 the impulse for its own contraction. Horridge (1955) has found that in Geryonia radial responses 

 involving movements of the proboscis can take place at the same time as rhythmic pulsation of the 

 circular swimming muscle, which suggests that one or other of the responses is independent of the 

 nervous system. The histological study of Krasihska (1914) indicates that the nerve-plexus is connected 

 not with the circular muscle but with the radial. Thus, a variety of evidence points to the independent 

 myoid conduction of the rhythmic swimming impulses in certain, if not in all, Hydrozoa. In the 

 Scyphozoan medusae, where two nerve-plexuses are present (Horridge, 1956a), the situation is 

 probably different. 



Physalia, by virtue of its large size and sturdy construction, is a good object for histological study. 

 All other common siphonophores are extremely delicate, and it is with justice that Bolles Lee (1900) 

 wrote of them : ' This group contains some of the most difficult forms to preserve that are to be found 

 in the whole range of the animal kingdom.' The tendency toward fragmentation is often a severe 

 handicap to the investigator. In Physalia, the gonodendra tend to break off when the animal is 

 fixed, but otherwise the parts remain intact. 



In the account which follows, reference will be made to the work of previous authors where 

 appropriate, but it may be stated at this point that the basic work on Physalia is Chun's section in 

 Bronn's Thier-Reich (Chun, 1897, 1902). The present account is, in the main, supplementary to it. 



2. Histological techniques 

 Specimens of Physalia quickly deteriorate in the laboratory, unless the water in which they are kept 

 is well oxygenated. Only freshly caught specimens were used in this investigation. 



Pieces of tissue were removed, washed briefly in distilled water, and placed in one or other of the 

 following fixatives: Flemming's fluid without acetic acid ('F.W.A.'), Baker's formaldehyde-calcium 

 (Ca-formaldehyde), both made isotonic with the seawater (Pantin, 1948); Zenker's and Helly's Fluids 

 (Baker, 1950); Bouin's and Carnoy's Fluids (Pantin, op. cit.), formaldehyde-sublimate-seawater, being 

 a mixture of ten parts formaldehyde with ninety parts saturated mercuric chloride in seawater. 



If the float was to be fixed, the fixing-fluid was injected directly into the coelenteron between the 

 codon and saccus, the whole float being immersed in the fixative at the same time. The intact float can 

 be fixed with virtually no shrinkage or distortion by this method, so long as the air inside is not 

 allowed to escape. When injecting the fixative care is therefore needed to avoid puncturing the saccus. 

 The fixative was injected with a hypodermic syringe, or in the case of fixatives containing mercuric 

 chloride, with a fine glass pipette. 



Wherever possible, whole strips of material were examined. Paraffin sections are hard to make, 

 more subject to distortion and altogether less revealing than thin strips of intact epithelium. In 

 certain cases, as in the study of the nervous system, sections were found to be almost useless. The 

 strips were prepared from fixed material under a binocular dissecting microscope, the tissue being 

 lightly stained beforehand in some dye, such as gentian violet, which could be washed out afterwards 

 in alcohol. The special staining techniques will be referred to where particular tissues are described. 

 Unless otherwise stated, Heidenhain's iron haematoxylin (Pantin, 1948) was used. This method was 

 found to be most generally useful, giving incomparably the best results with Flemming material, as 

 other students of the Coelenterata have found. 



