674 PROCEEDINGS OF THE ACADEMY OF [Oct., 



strongly shaded with purplish-blue almost to their tips. At the 

 base is an area of greenish-blue of greater or less extent, some- 

 times suffusing the inner web nearly to the tip, and in other exam- 

 ples almost entirely concealed at the extreme base of the feather. 

 Compared with other material before me, these specimens agree well 

 with one from Siam, but differ markedly from a Philippine speci- 

 men in the much smaller bill and in color pattern, the latter 

 matching Sharpe's figure of orientalis exactly. Sharpe credits 

 both forms to Borneo, calomjx being a winter visitor, and it is pos- 

 sible this is also the case in Sumatra. 



Pelargopsis capensis (Linn.). 



Alcedo capensis Linnajus, Syst. Nat. (XII), 1766, p. 180. "Cape of 

 Good Hope " prob.=Java, 



Four specimens from Larapong. 



The unfortunate complication of names which exists in this group 

 has given rise to a great diversity of treatment by different authors. 

 The facts in case are briefly as follows: Linnaeus (1755) gave the 

 name capensis to a kingfisher described by Brisson from the Cape 

 of Good Hope, but which we now know must have come from the 

 Malay region. 



Boddaert (1783) proposed the name Javana for the bird figured 

 by Daubenton {PL EnL, 757) as from Java, but which we now 

 know is the Bornean species. 



Gmelin (1788) based a name leucocephala on the same plate. 



Pearson (1841) described a species gurial from Bengal. 



In the light of our present knowledge it is possible to fix the 

 name javana upon the bird from Borneo (cf. Hartert, Nov. ZooL, 

 IX, p. 202), and the other two names upon races of a closel}'- 

 allied group found from India to Java. The name gurial with a 

 definite type locality is easily disposed of, which leaves capensis 

 Linn, for one of the remaining races. 



Sharpe (1870) was the first to designate these by name, pro- 

 posing floresiana for the bird from Flores, burmanica for the 

 Burmah foi'm, and malaccensis for that of the Malay peninsula. 

 At the same time he states that he considers Brisson's description, 

 upon which Linnoeus' name capensis was mainly based, to apply 

 to the Javan bird, but that Daubenton' s plate, al^o quoted by 

 Linnseus, represents his floresiana. Subsequently ( Cat. Bds. Brit. 

 Mm., XVll, p. 106) he considers Daubenton' s plate to have been 



