86 BULLETIN 61^ UNITED STATES NATIONAL MUSEUM. 



Table of scutcllation. 



Definite conclusions can not be based upon such a small number 

 of specimens, but, grantino; the accuracy of the locality, it should, I 

 believe, be concluded that these specimens are more than usually 

 dwarfed specimens of radix, and that the lateral stripe has become 

 slightly displaced in the reduction in the number of scale rows below 

 21. It is true that the writer has seen little evidence that the lateral 

 ^stripe tends to be disturbed in position when the fourth row is lost (see 

 also pp. 36-37), but in at least one specimen (Cat. No. 525, U. S. N. M., 

 from Racine, Wisconsin), which has 19-21-19-17 rows, the lateral 

 stripe apparently tends to descend upon the second row, where the 

 fourth row is dropped to leave less than 21 . This point should be care- 

 fully tested by the examination of a large scries of specimens from 

 eastern Wisconsin. 



Affinities. — If the lateral stripe is a safe index there can be no 

 uncertainty as to the inclusion of radix in the group to which we have 

 given its name, as there is no doubt as to the position of the lateral 

 stripe on the third and fourth scale rows. Just what its relation is to 

 marcianus is undetermined at present, but that it is not distantly 

 related to megalops is quite evident both from the position of the 

 lateral stripe and general coloration. Cope, in his assertion that 

 megalops is the representative of radix in Mexico, apparently recog- 

 nized the similarity of the forms, as did also Brown in 1904 (1904, 

 471), when he stated that " E. radix is a connecting link on the one 

 hand with E. proxima, * * * ^nd on the other in the southwest 

 with E. megalops." 



If, as I believe, marcianus is a member of the Radix group, it is 

 not likely that radix and megalops are directl}^ related, and this 

 improbability is enforced by the gap between their respective 

 ranges. It seems more probal)le that marcianus constitutes a link 

 between radix and megalops, and the closeness with which its range 

 coincides with the region between these two forms lends support 

 to this view. At any rate, radix can not at the present time be 

 connected with any other form in the same or adjacent regions, 

 with the exception of hutleri, which will be discussed later. It 

 differs constantly from sirfalis, eques, and elegans in the position of 

 the lateral stripe, and' from proximus and sauritus in the same region 

 by the increased number of scale rows and the persistently shorter 



