10 THE CYPRINODONTS. 



less modification, in others ; and in Heterandria, Glaridodon, and Girardinus, 

 there are three. In addition, on the stays, lateral processes are prominent 

 in some ; while in others tlie inferior stay alone is to be discovered. On 

 Xiphophorus there are four or five of the stays. Another feature of the 

 modification occurs in the males of some species, of Glaridodon for example, 

 in which the basal spines, to which the anterior anal rays are articulated, are' 

 much broadened. This is most pronounced in forms on which the clasper is 

 lon^-est, and it furnishes a broader base of attachment for the muscles con- 

 trolling the movements of the organ. 



By recent discussion attention has been directed to a decrease in the 

 number of vertebrae, of fishes in general, in and toward the torrid zone, and 

 several theories have been propounded to account for the phenomenon. The 

 species of this family, and others, have been somewhat carefully studied, — 

 first, to determine the facts, and, second, to test the theories. It is found 

 that those species making most use of the vertebra and the column have the 

 greatest number of vertebrae whatever the temperature, as may be seen in 

 the different genera together inhabiting Cuban waters, or those of Brazil, or 

 in the different species of a single genus like Orestias in Lake Titlcaca. It 

 is true a decrease obtains, with few exceptions, in the direction of warmer 

 waters, but warmth of water in such instances is attended by both increase 

 in the amount of food and decrease in the need of it, thus lessening the 

 comparative activity of the species. Some would ascribe the differences 

 directly to natural selection. This hypothesis of course cannot be proved ; 

 it begs the entire question. It is also found that with the decrease in the 

 number of vertebra;, there is in some cases a decrease in the number of fin 

 rays and scales. 



The stomach is an enlargement of the intestine ; it is not particularly 

 distinct, and is without a pouch or caecal appendages. As with higher 

 animals, the intestine is shorter in proportion to the greater amount of 

 flesh in the food. Variation in regard to feeding habits in the species of 

 a o-enus are readily indicated by comparisons of this organ. Fundulus in 

 o-eneral has a comparatively short intestine ; but in F. Kansae the tract is 

 much elongated. Lebias has a medium length, while its nearest ally, 

 Cyprinodon, has an intestine considerably longer. This organ in the mud- 

 eating species is very long. 



A membranous air bladder is present in all the known species of this 



