52 The Mechanism of Evolution" in Leptinotaesa 



Recureent Mutant L. multit^niata-intebmedia-meianothokax. 



This recurrent mutant has only been observed at Octlan in Jalisco, San 

 Bartolo in San Luis Potosi, and Sta. Eosalia in Chihuahua, It differs from the 

 same mutant form in the southern portion of the range of L. miiltitceniata, by its 

 smaller size and duller color of the head and puncture, which have little or no 

 luster. I have not been able to find this mutant in New Mexico and have no 

 knowledge of the relations of the " melanothorax " species found by Snow at 

 Las Vegas. These, while undoubtedly this same mutant type, do not seem to be 

 regular in occurrence in New Mexico, and in fact the mutant is rare over the 

 entire range of the habitudinal variety intermedia. (Plate 12, fig. 7.) 



Habitat, Ecology, and Life History. 



Like the preceding, this form passes the long, dry season (winter) in hiber- 

 nation as an adult in the ground. It emerges over most of its geographic 

 range when the summer rains begin, which is a variable climatic event. It 

 breeds at once, producing progeny that grow rapidly and give adults in 3 to 

 5 weeks ; these, over most of the range, after feeding for a few days may hiber- 

 nate or breed again, giving a second generation which hibernates soon after 

 emerging. Whether there be one or two generations per season is entirely 

 dependent upon the climatic conditions of the locality. In localities like Tuc- 

 son, where the summer rainy season may be short and not begin until August, 

 one generation seems to be the rule, although I found two in 1909. On the 

 plateau of Mexico, at Guadalupe, Zacatecas, two generations are the rule, 

 although here there are exceptions. When brought to the laboratory two gen- 

 erations are the normal cycle, but the second one of these is easily suppressed 

 by the use of desert complexes. Materials from nature are usually heterozygous 

 with respect to the reproductive rhythm and must be purified before they are 

 used in experiment. 



Ecologically this form presents nothing different in principle or in the 

 factors involved from L. multitceniata Stal, but differs considerably in minor 

 details and relationships, especially in the strictly desert areas. 



SOUECE OF MaTEEIAL. 



stock used for experimentation has been obtained at La Barca in Jalisco; 

 San Luis Potosi, Catorce, Monclova in Coahuila ; Jumilco, Zacatecas in Zacate- 

 cas; Chihuahua in Chihuahua, Mexico; Tucson in Arizona, and Deming in 

 New Mexico, and has been fairly imiform in composition from all locations. 



Habitudinal vabiety multit^niata-vaeiabilis nov. vab. 

 (Plate 4, figs. 10, 11.) 



In the upper portion of the Eio Lerma Valley, especially on the plains at the 

 foot of the Nevada de Toluca and eastward to the Sierra de las Cruces, occurs a 

 geographically isolated portion of L. muUitceniata, which differs therefrom in 

 constant genetically reproduced characteristics, and even though no fast line of 

 demarcation can be drawn, the gametic distinctness and the behavior of the 

 restricted group leaves one with little choice as to its taxonomic place. I have 



