70 BULLETIN 61, UNITED STATES NATIONAL MUSEUM. 



as a modification of tlic third and fourth row position, there is httle 

 difhculty in placing marcianus in this group. A prime difficulty 

 which presents itself is the necessity of accounting for the fact that 

 it overlaps in part the range of megalojys, with which, if we can trust 

 the evidence of the lateral stripe, it must be closely related. Before 

 this objection can be considered, it will be necessary to obtain much 

 more information on the distribution of both marcianus and megalops 

 in southern Arizona, for, as has been shown, but very few records of 

 either form are at hand from this region, while some of those that are 

 available are, to say the least, open to question. However, it must 

 be admitted that marcianus undoubtedly occurs at Tucson, while 

 megalops exists in southern New Mexico (Duck Creek), so that the 

 forms unquestionably overlap. " Still, it seems to us that the simi- 

 larities are close enough to warrant the working hypothesis that 

 marcianus is an offshoot of megalops that pushed across the deserts 

 of northeastern Mexico and into the Trans-Pecos and western Texas 

 region, and here obtained its individuality, so that as it moved east- 

 ward it found its range limited by the transition line between the 

 prairies and the forest, while in its subsequent w^estward movement 

 into the range of the parent stock {megalops) its differentiation 

 in structure or habits was sufficient to keep the two from inter- 

 breeding. To the northward, in Arizona and New Mexico, its range 

 was limited (except in the valleys) by the high plateaus, but in Texas 

 the extension of the semiarid prairie-plains furnished a highway to 

 the northward, along which it spread, becoming (in Oklahoma and 

 southern Kansas) reduced in the number of scale rows and darker in 

 color to constitute the form now known as radix. 



This is a bold hypothesis to be made on the basis of such a small 

 number of specimens, but, granting the fact that more observations 

 may overthrow it entirely, it seems to me, in the light of our present 

 knowledge, to be the most satisfactory explanation for the origin of 

 the form. At all events, if it arouses discussion and stimulates fur- 

 ther investigation, it will have served its purpose. 



RADIX, a 



Description. — ^I have already noted that along its northern boun- 

 dary, in Oklahoma, marcianus approaches another form, radix, with 

 wliich it is apparently closely related. However this may be, the 

 specimens now at hand indicate that throughout most of its range 

 radix is a distinct and well-defined form, which it is comparatively 

 easy to distinguish. 



o Thamnophis radix (Baird and Girard), Catalogue of North American Reptiles, 

 1853, p. 34. Includes Eutaenia haydeni Kennicott, E. radix twiningi Coues and 

 Yarrow, and E. radix vielanotaenia Cope. 



