Fishes of the Western North Atlantic 9 



in their body fluids by a complicated excretory process. Thus, via their gill membranes, 

 they excrete most of the salt from the sea water that is consumed, and at the same time 

 they reduce their excretion of water to the minimum that will provide the necessary 

 amount of urine. In some groups this reduction is effected by atrophy or loss of the 

 glomeruli,-* but in other saltwater groups that still retain the glomeruli, including the 

 living coelacanths (jJ: 2588, Millot and Anthony), this goal appears to be reached by 

 constriction of the glomerular arterioles.-* 



We have yet to learn how bony fishes that spend part of their life in fresh water 

 and part in salt water or that suffer no apparent harm from rapid changes in the 

 salinity of the water manage to maintain themselves in osmotic balance with their sur- 

 roundings. This applies equally to such of the cyclostomes as spawn and pass through 

 their early stages in fresh water but make their growth in salt water. Those that pass 

 their entire life in salt water are isotonic to their surroundings, or very nearly so (see 

 especially McFarland and Munz, 4.8: 348). 



Subclasses. Students of phylogeny have disagreed widely concerning the relation- 

 ships of the various groups that make up the ichthyological tree. In the first instance, 

 a decision on the best arrangement of the various groups of the Class Osteichthyes 

 into Subclasses depends on the relationship of the Dipnoi (lungfishes) to the other 

 bony fishes. Johannes Miiller (5j), writing in 1844, rated the Dipnoi as a separate 

 Subclass of the Class Pisces, and this scheme was later adopted by Smith Woodward 

 {gi: 234), by Goodrich (.^5: 230), by Bridge, and by Jordan (jp). Berg's (4) later 

 characterization of them as a class of the Series Pisces embodies this same point of view, 

 as does Bertin and Arambourg's treatment of them as the Subclass Dipneusti of the 

 Class Osteichthyes (jJ: 2522). It is likely, indeed, that this scheme would still be 

 adopted generally were our knowledge of the Dipnoi limited to those of today, for these 

 are set apart from all other living groups of bony fishes by the solid fusion of the palato- 

 pterygoid-quadrate elements of the upper jaw-complex with the lower surface of the 

 skull, by the general morphology of their brains and hearts, and by their total lack of 

 both premaxillary and maxillary bones (a significant difference); in addition, the living 

 Dipnoi are set apart further from all other bony fishes except the coelacanths by the 

 presence of a jointed axial skeleton in their paired fins (Fig. 2g). None of the other 

 various features that have been cited as alternative between Dipnoi and other bony fishes 

 is strictlv so. Palaeontological evidence, furthermore, is to the effect that the oldest 

 known fossil Dipnoi (Devonian) resemble in many characters the Rhipidistia (Middle 

 Devonian), whose nearest relatives among present-day fishes are the coelacanths 



{73- 114)- 



Accordingly, the Dipnoi and the coelacanths among living fishes have been 

 united by several recent writers {55: 376; yi [1937]: 327; 75: 589; yy. 227; 4y: 



28. Examples cited by Smith [jg: 104) are Lophitts (anglerfishes), Opsanus (toadfishes), Hippocampus (seahorses), and 

 Syngnathus (pipefishes). 



29. For more extended discussions of this general subject, from which the foregoing summary has been drawn, see 

 Prosser et al. (6^: 48-52), Smith (79: 62-64, i02-t05), Romer (74: 392-393), Denison (20: 429-430, 439-441), 

 and Black {12: 182-199, ^^^^- 199-206). 



