Bidentogon Buckett and Gardner, 1968

Shelley, Rowland M., 2003, A new polydesmid milliped genus and two new species from Oregon and Washington, U. S. A., with a review of Bidentogon Buckett and Gardner, 1968, and a summary of the family in Western North America (Polydesmida: Polydesmidae), Zootaxa 296 (1), pp. 1-12 : 9-10

Genus Bidentogon Buckett and Gardner, 1968

Bidentogon Buckett and Gardner, 1968:198 . Shear, 1972:489­490. Hoffman, 1980:172; 1999:438. Shelley et al., 2000:87–88. Shelley, 2002:108.

Type species. B. helferorum Buckett and Gardner, 1968 , by original designation.

Diagnosis. A genus of minute Polydesmidae without distinct pulvillus but with scattered hairs along caudal surface of gonopodal telopodite; division point located distad, tibiotarsus and endomerite widely separated, subparallel to each other.

Species. Two

Distribution. The coast of California from Mendocino to Santa Cruz counties, extending inland to Sacramento County ( Fig. 1 [5]).

Remarks. Chamberlin (1918) based his description of Brachydesmus californicus on a sample from Stanford, Santa Clara County, containing only one male, which is therefore fixed as the holotype by monotypy (Article 73.1.2 of the Code); the date of collection was not provided. During the next 46 years, Attems (1940), Chamberlin & Hoffman (1958), and Buckett (1964) all recorded Stanford as the type locality, but Shear (1972) reported discovering the type sample at the MCZ, which contained an unspecified number of male and female syntypes, was collected in January 1914 from Sacramento, Sacramento County, and was labeled “ ‘ HOLOTYPE.’ ” Subsequently, Hoffman (1999) and Shelley (2002) reported Sacramento as the type locality. I have examined this sample and found it to contain two males and three females with a label stating “ Brachydesmus californicus sp. n. ” instead of one stating “ HOLOTYPE.” Evidently, Chamberlin or somebody else mistakenly placed the “ sp. n.” label with the Sacramento sample because it does not match the locality or number of specimens reported by Chamberlin (1918) in the original description. There is a vial with one male at the NMNH that was collected at Stanford in 1910 and matches the requirements for the type, but the possibility cannot be discounted that there might have been another pre­1918 sample from Stanford that is now lost. As both the Sacramento sample and that taken at Stanford in 1910 were collected prior to Chamberlin’s description (1918) of B. californicus , one can conjecture that he had both samples on hand when preparing the account and placed the label in the wrong vial himself. We will never know what happened, but Chamberlin’s clear intent, as evidenced by the published account, was to apply this name to the form at Stanford, the designated type locality. I have therefore placed a label in this vial stating “ Brachydesmus californicus Chamberlin, 1918 , PROBABLE HOLOTYPE " and deposited it in the NMNH type collection.

Shear (1972) noted that the endomerite in the males from Sacramento is apically broad whereas that of the type of Bidentogon helferorum , from 1 mi (1.6 km) north of Mendocino, Mendocino County, is narrow and acuminate. I have examined most of the available males of Bidentogon and found that all those collected along the coast and San Francisco Bay, including Stanford, exhibit the latter condition; the only sample in which the tibiotarsus is broad is that from Sacramento. Consequently, as the true type localities of both nominal species occur along the coast, its name must be changed to californicus , which holds priority over helferorum by 50 years. No name is available for that from Sacramento, so I propose Bidentogon expansus n. sp. I have placed the type label for B. expansus in this vial along with one stating that Stanford is the true type locality for Brachydesmus californicus .

Figures

FIGURE 1. Occurrence of the Polydesmidae in western North America; 1, Scytonotus; 2, Calianotus; 3, Utadesmus; 4, Speodesmus; 5, Bidentogon; 6, Retrorsia.