structure and systematics s9 



The larval swimbladder 

 The teleost swimbladder arises early in development as an outgrowth of the dorsal or lateral walls of 

 the foregut. While the embryonic formation of the swimbladder in a deep-sea fish has yet to be 

 described, the sequence of events is likely to be very similar to that in shallow water fishes. When 

 considering the general features of myctophid larvae, Holt and Byrne (191 1) mentioned that from a 

 very early stage the roof of the swimbladder appeared to be darkly pigmented. It is also clear from 

 Jespersen and Taning's (1926) figures that the early post-larvae of gonostomatid fishes have a well- 

 formed swimbladder. In post-larval Vinciguerria and Maurolicus I found the swimbladder to be 

 full of gas. 



The presence of gas in the larval swimbladder raises an interesting question concerning the early 

 functioning of the organ in the deep-sea environment. In many shallow water fishes the sac is first 

 inflated by the larvae gulping in air at the surface and passing it down the larval pneumatic duct 

 (which disappears during later development in physoclists). While the larval life of most bathypelagic 

 fishes is passed in the surface-layers, it seems likely that the eggs are shed at deeper levels. Hatching 

 may thus take place as the eggs float upwards from the depths. If so, many larvae could be far from 

 the surface-film when the larval swimbladder is ready to be filled with gas. 



However, McEwan (1940) found that, in Hemichromis, the connection between the larval swim- 

 bladder and the gut never developed a lumen and, to make sure of the implications of this discovery, 

 the early larvae were denied experimentally all access to the surface. In spite of this the larval swim- 

 bladder became filled with gas. McEwen found that at one stage the lumen was obliterated by the 

 swelling of highly vacuolated cells forming the inner epithelium. Having expanded to the limit, these 

 cells suddenly collapsed to form a flat epithelium, after which gas appeared in the cavity. If such 

 a mechanism is found in deep-sea fishes, it is obvious that the larvae need not seek the surface- 

 film in order to initiate the use of the swimbladder as a hydrostatic organ. 



At least five species of Cyclothone (braueri, signata, microdon, pygmaea, and acclinidens) must have 

 a gas-filled swimbladder during their larval life, but after metamorphosis the organ regresses and 

 becomes invested with fatty tissue. Presumably this is also true for other deep-water fishes with 

 a fat-invested swimbladder in the adult phase. All such species will best be considered in a 

 separate section (pp. 65-68). In the Miripinnati (Bertelsen and Marshall, 1956) and Stylophorus, 

 the larval swimbladder also undergoes regression, but does not serve as an attachment for fat in 

 the adult. 



Lastly, all available evidence shows that bathypelagic fishes, without any trace of a swimbladder 

 when adult, are also without a definite larval organ. (I have examined larval paralepidids, scopel- 

 archids, and melanostomiatids without finding a swimbladder.) Furthermore, Bertelsen (1951) 

 remarked that larval ceratioids, like the adults, have no swimbladder, but he suggested that 'gela- 

 tinous tissue under the skin, which is present in all Ceratioid larvae in more or less well-developed 

 condition, could be regarded as a floating organ '. More recently, Shelbourne (1956) has seen a corre- 

 lation between the pelagic habit in fish eggs and larvae and the development of large subdermal 

 spaces in the young stages. As these spaces seem to be filled with low density fluids derived from the 

 yolk, they act as buoyancy chambers. It is clear from Shelbourne's figures that voluminous subdermal 

 spaces are developed regardless of the presence or absence of a swimbladder in the larval phase. 

 This is also true of bathypelagic fishes. Larval myctophids belonging to species with a well-formed 

 adult swimbladder (e.g. Benthosema glaciale, Hygophum benoiti, Myctophum punctatum) have large 

 subdermal spaces, particularly in the head region (see Taning's (191 8) figures). Regarding fishes 

 without a swimbladder, we have already referred to Bertelsen's findings in the ceratioids. It would 



8-2 



WOODS 



HOLE, 



MASS. 



