M. P. Gervais on the Freshwater Fishes of Algei^ia. 137 



marine species to which the generic name Glyphisodon has been 

 given ; and yet there are no species of Gbjphisodon in the 

 Mediterranean sea. This species must finally be compared with 

 the numerous Pharyngognathi, often taken for Chromides, which 

 MM. Natterer and De Castelnau* have brought from the great 

 rivers of intertropical America, although we may be sure, by the 

 knowledge of its maxillary teeth, and the comparison of its 

 pharyngeal teeth with those of the species taken by M. Heckelf 

 as the types of his difFereat genera, that Sparus Desfontainii has 

 more analogy with the Glyphisodonts than with the best-known 

 American species. 



In the reflections with which Mr. Tristram has accompanied 

 the description of the species, probably identical with the Bolti, 

 to which Dr. Giinther has given his name, this naturalist inquires 

 whether the fish which he brought from Tuggurth is not to be 

 regarded as a last living vestige of the fauna which peopled the 

 Saharan sea during the Tertiary epoch, '^ before,^' as he says, 

 "the elevation of the soil of North Africa poured into the 

 Mediterranean the waters of this vanished ocean." I do not 

 know whether the Sparus Desfontainii^ a fish which approaches 

 marine species of fishes still more closely than the Bolti, Tilapia, 

 Coptodus^ ov Haligene, can serve as an argument in favour of this 

 opinion. Before according any credit to such an hypothesis as 

 this, I should like to know that the same species of Fish had been 

 observed in the marine strata deposited during the period to 

 which this supposition would make it ascend ; and this has not 

 beeu done; but, as regards Haligenes Tiistrami and its com- 

 panion in its habitat (and, no doubt, synonym), Coptodus Zillii, 

 I cannot avoid remarking how much their identification with the 

 Bolti or Chromis of the Nile and other fresh waters of Africa, 

 is opposed to this supposition. 



Nor do I think we have any more reason for adopting the 

 opinion which has sometimes been put forward with regard to 

 Cyprinodonts thrown out, like the Coptodons, by artesian wells 

 of the Sahara, namely, that they are derived from a sea extend- 

 ing beneath this region — since, wherever they are met with, the 

 Cyprinodons, like the Boltis themselves, are non-marine fishes, 

 whether they have been taken in Algeria, Portugal, Spain, Syria, 

 Egypt, Abyssinia, or even in America. 



If we wished to find in geological time an equivalent to the 

 ichthyological fauna of Algeria, which is related to the fluviatile 

 faunas of Southern Europe or of the rest of Africa by the whole 

 of its known genera, it is in the lacustrine deposits of the 



* Animaux nouveaux ou rares do TAm^rique du Sud. 

 t Annales du Mus^e de Vienne, tome ii. 



Ann. S^ Mag. N. Hist. Ser. 3. Vol.xix. 10 



