176 VERTEBRATA. 



of them, it is not easy to determine their exact number. The head 

 contains the heart and breathing-organs as well as the brain. 



The organ of hearing (it has been doubted if they hear at all) 

 is rather complicated in fishes, although a tympanum and cochlea 

 are absent ; two osseous bodies (otolites) are generally found in 

 the vestibule. 



The teeth are not always present, e.g. pipe-fishes, sturgeons, &c. 

 In others they are very numerous, and attached to other bones 

 besides maxillae and mandibles ; they are shed and renewed 

 throughout the whole of their lives. 



In their earlier stages of growth many fishes are known to un- 

 dergo great changes, nor is growth known to be definitely ar- 

 rested at maturity. The eye, however, ceases to grow at an early 

 stage, so that "old fishes have comparatively smaller eyes than 

 young ones." Fishes are very prolific ; the roe of the cod is said 

 to contain nine millions of ova. 



A few species are viviparous. In no fish is there a trace of an 

 amnion or of an allantois. 



Fish swimming near the surface respire more oxygen than those 

 living at greater depths ; hence they die soon after they are taken 

 out of the water, such are pilchards, salmon, mackerel, &c. Many 

 are enabled to live in mud when hardened and dried up by the 

 sun, and others may be kept for a time in a frozen state without 

 destroying life. 



Among the many deep-sea fishes discovered during the ' Chal- 

 lenger' expedition, one, Halosaurus restrains, was found to live 

 at a depth of 2750 fathoms. 



Fishes were most abundant during the Old Red Sandstone 

 epoch ; the earliest traces of them are found in the Silurian rocks. 

 Of the recent species some 12,000 are described ; of these only 

 about 220 inhabit the English seas and rivers, but there are no 

 seas which yield so many for the table, whether in species or in- 

 dividuals. 



It is difficult to find two writers agreeing upon the classification 

 of fishes. Some consider that the class should be so limited as to 

 exclude the Pharyngobranchii, the Marsipobranchii, and the Dip- 

 noi, each of these also forming a class. Dr. Gtinther, taking 

 the three groups as subclasses, and uniting Chondropterygii to 

 the Ganoidei under the name of Palaeichthyes, makes, together 

 with the Teleostei, five subclasses. Prof. Huxley more recently 

 adheres to the six orders adopted by him some years ago, though 

 he has since proposed certain primary divisions in reference prin- 

 cipally to the mode of attachment of the jaws to the skull. 

 Another classification, founded almost exclusively on the skeleton, 



