ATOPETHOLID MILLIPEDS — HOFFMAN AND ORCUTT 127 



Six ocellus counts are as follows: 33, 35, 37, 37, 39, and 42. Here 

 there seems to be neither sexual nor geographic variation, but the 

 series is too short to be of particular value. 



The size appears to vary considerably. The largest known speci- 

 men is a female from Garcia, Neuvo Le6n, 75 mm. long and 8.0 mm. 

 wide; the smallest is a female from Pandale, Texas, about 40 mm. 

 long and 5.0 mm. wide. The small specimens appear to be struc- 

 turally identical with larger ones. 



Clypeal foveolae vary from 4-4 to 5-5, and are somewhat more 

 constant than in E. soleatus, in which counts of 3-3 are noted as well. 



Synonymy: The synonjTny of Eurelus kerrensis and Toltecolus 

 garcianus was established in the following manner. The adult male 

 from Real County, Texas, was found to agree perfectly with the 

 drawings given for garcianus as well as with the description, as far 

 as it goes. This specimen also matched the original diagnosis of 

 Cenirelus very closely, leaving little doubt that it is congeneric with 

 C. falcatus, and on this basis we determined two collections from 

 Texas as Centrelus garcianus. However, when a map of the known 

 records of the Eurelinae was prepared, it was observed that Real 

 County is geographically adjacent to Kerr County, whence came the 

 type series of E. kerrensis. Comparison of our material with the 

 description of that species, moreover, showed a remarkable con- 

 cordance in every particular except that metazonite spines were not 

 mentioned for kerrensis. On the chance that they may have been 

 overlooked, we requested Chamberlin to examine the types for this 

 character. He very kindly did and reported (in a letter dated May 

 30, 1957) that small but distinct and upturned spines are actually 

 present. This discovery removed the last obstacle for presuming 

 that kerrensis is the older of two names applied to a spiroboloid 

 population which occurs in the plateau country of Texas and adjacent 

 Nuevo Le6n. Although we have been unable to study the type 

 specimens of either name, we feel that the available evidence as 

 outlined above makes the likelihood of error quite remote. 



Distribution: Centrelus kerrensis is now known to have a fairly 

 wide range over much of western Texas and adjacent Nuevo Le6n. 

 From the existing records and the fact that it has not so far been found 

 in the Rio Grande Valley, we assume that it is an upland form. The 

 type locality was Raven Ranch, in Kerr County, Texas (Chamberlin 

 and Mulaik, 1941), and the species was subsequently recorded (under 

 the name Toltecolus garcianus) from Garcia, near Monterey, Nuevo 

 Le6n (Chamberlin, 1943a). Material has been seen from the following 

 new localities: 



Texas: Real County: 10 miles north of Leakey, June 10; 1955, Leslie Hubricht 

 (R. L. Hoflf man) . Presidio County : Near Porvenir, September 28, 29, 1946, Bryan 

 511799—60 3 



