39 o DISCOVERY REPORTS 



(PI. XXVII, fig. 6). The reaction appears to take place both in the granules and in the matter interspersed 

 among them. The secretion generally shows considerable tenacity towards haematoxylin, as noted by 

 Berrill (1949) in the case of similar cells in Obelia. There appears to be little if any metachromasia 

 with toluidin blue. In some gelatin sections of Ca-formaldehyde material a diffuse background 

 coloration was obtained with Sudan black. This is of interest in view of the finding that a lipid-like 

 material occurs in the cuticle. 



The gland-cells present a fairly uniform appearance throughout the animal and it has not been 

 possible to distinguish more than one type of cell. It is of interest in this connection to note the 

 discovery by Mettey and Hamon (1949) of two types of gland-cell in the ectoderm of Abylopsis 

 tetragona (Calycophora). No hint is given of their functional significance, but one of them ('cellule 

 a substance hyaline') has histological features in common with the gland-cells in Physalia. In 

 Physalia, if secretion of substances other than those of the cuticle were involved, mucus would merit 

 consideration. Specimens kept in captivity sometimes produce a glutinous substance, but this seems 

 to come from the endodermal gland-cells of the gastrozooids. 



Not only in the Siphonophora, but in the Hydrozoa generally, cuticle secretion presents many 

 problems. Bonner (1955) found that, in the planula of Phialidium, gland cells giving a positive P.A.S. 

 reaction are concentrated in the anterior pole, where a chitinous secretion appears at the time when the 

 larva fixes itself. As with Physalia, the chitinous secretion has not been definitely traced to the gland- 

 cells. Bonner further found that the gland-cells disappear after fixation, although the perisarc 

 continues to form around the column of the young hydroid. Although this might suggest that the 

 gland-cells are only needed when a copious amount of chitin is to be produced (that is, for attach- 

 ment), it might also mean that the cells produce some substance which mixes with the chitin at the time 

 of attachment but is not produced at other times. In the actinula larva of Tubularia larynx Pyefinch 

 and Downing (1 949) found evidence for an ' extra-chitinous cement ' substance in the attachment region. 

 However, a cuticular sheath is also present in this region, and Ciamician (1879) reported that in 

 T. mesembryanthemum such a sheath (early perisarc) is produced by gland cells in the aboral ectoderm. 



Thus, although there is a possibility that the gland-cells, which are found in the aboral ectoderm of 

 planula and actinula larvae, produce something other than chitin, this is unlikely. Bonner's description 

 of the gland-cells in Phialidium accords with the present writer's findings on Physalia ; in both cases 

 the major part of the evidence points toward chitin as the substance secreted. 



6. The gas-gland 



The gas-gland occupies a circular thickened area on the lower surface of the pneumatosaccus, toward 

 the side b-c (Text-fig. 1) of the float. 



The length, breadth and depth of the float may vary considerably according to the posture of the 

 animal, but the diameter of the gas-gland (g) is less variable, and can be used as a rough index of the 

 size (and, perhaps, age) of the specimen. Collapse of the float leads to very little alteration in g. It 

 would be of interest to calibrate degrees of morphological complexity in the budding pattern against 

 g values. 



According to Haeckel (1888) the gas-gland of the largest specimens of Physalia may measure 

 10-15 cm - across. In a large 'Discovery' specimen in the University Museum, Oxford, g = 5'2$ cm.; 

 in the largest 'Discovery' specimen in the British Museum, < §'=4 , o cm. Compared with these, the 

 specimens collected in the Canary Islands were small: in the largest £ = 2-5 cm., and, in the smallest, 

 £ = 0-35 cm. 



The histology of the gas-gland in Physalia has been studied by Dahlgren and Kepner (1908). 

 A section through the gas-gland of a very young specimen is given by Okada (reproduced by Hyman, 



