MILLIPEDS — HOFFMAN 195 



Distribution: P. clarus, the only member of the genus known 

 from west of the Mississippi River, is endemic to the upland parishes 

 of western Louisiana above the Kisatchi escarpment. A single Texan 

 locality is known, and it seems likely that the species will eventually 

 be discovered in southern Arkansas. Material has been seen from 

 the following localities: 



Louisiana. Grant Parish: west of Pollock, 3 9 ? , July 9, 1955, Leslie 

 Hubricht (RLH); Dry Prong, 2 9 9, May 9, 1954, H. S. Dybas (CNHM). 

 LaSalle Parish: Jena, 4 d'cf, 5 9 9, Feb. 28, 1954, J. Stone (NBC). Lincoln 

 Parish: Ruston, 1 c?", 1 9 , Feb. 15, 1952, W. J.Harmon (NBC). Natchitoches 

 Parish: Creston, 3 cf' c?", 6 9 9 , types, February-March 1915, K. P. Schmidt 

 (MCZ, USNM) ; Bellwood, 1 9 , Oct. 3, 1953, J. D. Montgomery (NBC) ; Chastine, 

 2 miles north of Creston, 1 cf, 3 9 9, Apr. 29, 1947, J. M. Schmidt (CNHM). 

 Rapides Parish: Forest Hill, 2 9 9, Nov. 11, 1945, R. L. Wenzel (CNHM). 



Texas. Newton County: Weirgate, 1 9 , Mar. 28, 1943, Barth A. Maina 

 (CNHM). 



This species has also been recorded && Pachydesmus kisatchinsis, 

 from Kisatchi, Natchitoches Parish, La. (1 cf , Mar. 29, 1937, Leslie 

 Hubricht), by Chamberlin (1942). 



Pachydestniis crassicutis (Wood) 



Under this name I propose to associate all of the large pachydesmids 

 known from the region east of the Mississippi River. In the past, 

 each has been considered a full species, but the evidence of morphology 

 and geographic distribution now available indicates that all of the 

 valid forms represent only geographic races of a single widespread 

 species. It has also become obvious, through direct comparison of the 

 different forms, that the structm-al peculiarities upon which the names 

 were based are much less real than apparent. 



Although there seems to be a great difference between such forms 

 as P. c. laticollis and P. c. incursus, other subspecies largely bridge the 

 gap between them, and there is actually a geographic gradation to be 

 observed in every character which might be considered diagnostic of 

 a species. 



No two forms of the crassicutis group occupy the same territory. 

 Instead, the ranges are now known to be entu*ely complimentary, 

 and one definite case of intergradation is known. Specimens taken 

 between the ranges of P. c. laticollis and P. c. retrorsus are clearly inter- 

 mediate as regards the structure of the gonopods. 



Despite the general picture of geographic variability, I believe that 

 most of the forms here considered valid will prove to be stable sub- 

 specific units. Except for minor, presumably individual variations, 

 specimens taken through the extent of the range of a given subspecies 

 are essentially alike, and there is no reason to believe that we are 



