REVISION OF THE KING SNAKES. 143 



Range. — This form is apparently the only representative of the 

 genus in Central America. It is known from Costa Rica to southern 

 Mexico; in the latter country it occurs at least as far north on the 

 east side as Tuxpan in the Province of Vera Cruz, and, on the Pacific 

 side, it is known from Oaxaca, Guerrero, and southern Michoacan. 

 According to Cope (1900, 900) Sumichrast says of it in his manu- 

 script notes: ''It is distributed throughout the warm and temperate 

 regions, but disappears in the alpine region, where, at least, I have 

 never observed it." Apparently the only record at variance with 

 this distribution is one for the ''City of Mexico" (Boulenger, 1894, 

 204), and this may easily be an error. 



Published records for other localities than those listed are as follows : 

 Haute Vera Paz (Salvin, 1861, 227; Bocourt, 1886, 614), Retalhuleu 

 (McLain, 1899, 5), and Coban, Guatemala (Werner, 1903, 347); 

 Teziutlan (State of Puebla), and Misantla (State of Vera Cruz) 

 (Ferrari-Perez, 1886, 187); Acapulco, Mexico (Garman, S., 1883, 67); 

 Oaxaca (Bocourt, 1886, 616); Carrizal, South Michoacan, west of the 

 lower Balsas River (Gadow, 1911, 6 and 24); Zamora, Michoacan 

 (Duges, 1876, 222); Chiipan-cingo, San Luis Allende (Gadow, 1905, 

 196); in Boulenger' s Catalogue of Snakes (1894, 204) are the follow- 

 ing records: Tierra Colorado and Amula, Guerrero; Teapa, Tabasco; 

 Huatuzco; Belize; Duenas, Guatemala; Irazu, Costa Rica; Chiriqui;. 

 City of Mexico. 



Variation. — The necessity of a knowledge of the variations in 

 a form is well illustrated by the present case. Without regard for 

 normal variation and geographic probabilities, numerous names 

 have been applied to specimens from Mexico and Central America 

 with the result that quite distinct forms have been lumped together 

 and individual variations of the same form, even from the same 

 locality, have been granted distinction. Confusion is the only end 

 attained, and that was the situation in the beginning. It is by no 

 means impossible that fuller series of specimens may demonstrate 

 the necessity for further subdivision, but it is hoped that a natural 

 group has been distinguished in polyzona as here defined, so that no 

 radical rearrangements will be necessary. 



The diagram of extremes and averages of ventral plates for certain 

 regions (fig. 44) shows that throughout all the central portion of the 

 range, from Nicaragua to southern Mexico, the numbers are high and 

 fairly constant. North through Tehuantepec and Vera Cruz the 

 decrease is distinct but not great. The material from Yucatan and 

 Guerrero is too scanty to be more than suggestive, but differentiation 

 by isolation is indicated in the former locality and either that or 

 approach to nelsoni in Michoacan is the inference in the latter. Costa 

 Rica and Panama show a definite passage to the low average of 

 micropholis. 



