NO. 2264. BEES IN THE NATIONAL MUSEUM. 3—COCKERELL. 183 



ANTHOPHORA ACERVORUM, variety SQUALENS Dour«. 



Soochow, China (N. Gist Gee, males No. 123, females No. 124). I 

 thought at first to separate the Chinese form as a distinct variety; 

 but after examining specimens of four different European varieties 

 there seemed to be no advantage in attempting to define another, 

 though the Chinese specimens are slightly different from the available 

 European ones. The males have a large yellow spot on mandibles, 

 and the middle tarsi have joints 2 to 4, and base of 5, red. 



ANTHOPHORA VIGILANS Smith. 



Buitenzorg, Java, March, 1909 (Bryant and Palmer) ; Tjibodas, 

 Mount Gede, Java, April 20, 1909 (Bryant and Palmer) ; Mount 

 Salak, Java, 2,500 feet, March 7, 1909 (Bryant and Pahner). The 

 male has very large foot-shaped black marks on clypeus. 



ANTHOPHORA TEXANA Cresson. 



Mexico, D. F. (J. K. Inda 39) ; Mexico (Baker coll. 2320). 



CROCISA DECORA Smith. 



Buitenzorg, Java, March (Bryant and Palmer) ; Depok, Java 

 (Bryant and Palmer) ; Mount Salak, Java, above 3,000 feet, May 4, 

 1909 (Bryant and Palmer). The Depok one has the markings pale 

 greenish blue, a different tint from that of the others, but it seems to 

 represent only an individual variation. 



CROCISA JAPONICA Frleie. 



Kuling, Kiangsi, China (N. Gist Gee). This is perhaps racially 

 separable from the Japanese insect, but I do not venture at present 

 to describe it as distinct. It is very close to O. kanshireana Cock- 

 erell from Formosa, but is easily separated by the blue spots on the 

 lobes of the scutellum. It is distinct from C. sv/rda Cockerell. 



CROCISA IRISANA HUMILIS, new ssbspecies. 



Buitenzorg, Java, March and April (Bryant and Palmer) ; Pela- 

 boean, Eatoe, Java (Bryant and Palmer). 



Type.— C^t. No. 20716, U.S.N.M. 



Length of female, 9 mm. 



Typical C. irisana is from the Philippine Islands ; the Javan form 

 has very dark wings, the tarsi with hardly any pale hair, and the 

 scutellar notch with very little or no pale hair above. The type- 

 locality of the subspecies is Buitenzorg. It is a rather poorly 

 defined form, but four specimens are before me and seem to indicate 

 that the Javan insect is recognizably distinct. Friese records this 

 from Java as C. pulchella Guerin, but the original pulchella (never 

 properly described) came from New Ireland and was supposed to be 

 a variety of C. rdtidula Fabricius. 



