378 THE NATURALIST IX NICARAGUA. [Ch. XXI. 



boundary between the two sub-provinces is not found, as 

 was supposed, in the depression of the isthmus occupied 

 by the great lakes and their outlet the San Juan river, 

 but must exist further towards, if not in, Honduras. 

 Mr. Salvin says, " What I suspect to be the case, though 

 I cannot as yet bring evidence to prove it, is, that the 

 forests of Chontales spread uninterruptedly into Costa 

 Rica, but that towards the north and north-west a 

 decided break occurs, and that this break determines the 

 range of the prevalent Costa Rican and Guatemalan 

 forest forms."* I can confirm Mr. Salvin' s supposition. 

 The San Juan river forms no greater break in the forest 

 than a dozen other rivers that run through it and fall 

 into the Atlantic. But a decided interruption does occur 

 to the north-west. It is found in the valleys of Humuya 

 and Goascoran in Honduras, which, along with the 

 central plain of Comayagua, constitute a great trans- 

 verse valley running north and south from sea to sea, 

 and cutting; completely through the chain of the corclil- 

 leras.f The highest point of this pass is 2850 feet above 

 the sea, and the country around is composed of undu- 

 lating savannahs and plains covered with grass. The 

 Gulf of Honduras, cutting deeply into the continent, 

 also plays an important part in preventing the inter- 

 mingling of the faunas of the two sub-provinces, but the 

 principal barrier is the termination of the great Atlantic 

 forest north-westward, which even at Cape Gracias begins 

 to give place to plains and savannahs next the coast, 

 Mv entomological collections were much more com- 



*> o 



plcte than my collections of birds, especially as to the 



* The Ibis," July, 1872, p. 312. 



| Squier, " States of Central America," p. G81. 



