( 54 ) 



The sexes seem not to differ in colour or size. 



Nests were found on Guam in February and Marcli, two to six feet above the 

 ground. They are constructed in the usual way of the Rhipidura nests, being a 

 round cup with a very smooth whitish outer surface, and rest on the top of a branch. 

 The eggs are creamy buff, with a zone of tiny patches and sjwts of brown and 

 greyish, sometimes nearer the thick end, sometimes right in the middle. They are 

 somewhat longisli, and measure about 16-o : l^-5 mm. 



8. Rhipidura saipanensis sp. uov. 



Rhipidura versicolor Oostalet I. pp. 190 ft", partim (nee Rh. rersicolor Hartlaub & Finsch). 



The specimens from Saipan are very similar to F/i. versicolor, from Yap in the 

 Mackenzie group, but differ in having the bases of all the rectrices rnfous, the rump 

 and upper tail-covcrts rufous. The sides of tlie abdomen are not olive-brown, but 

 rufous. These differences seem quite sufficient to give a name to the Saipan form, 

 especially since the distribution of a form over the Mackenzie group and Saipan, 

 while the island of Guam between those two areas is inhabited by lih. uraniae, 

 would be almost jihenomenal. The ear-coverts, lores, and a line under the eye are 

 black. The nnder tail-coverts are of the same pale rufous colour as the flanks, 

 while the rump, upper tail-coverts, and base of tail are of a deep bright rnfous. 

 The rnfous at the base of the rectrices, which on the central pair occupies nearly 

 half the feather, diminishes towards the sides of the tail, it e.xtending only 

 10 — 15 mm. wide on the lateral pair. The white, on the other hand, which forms 

 only a narrow fringe on the central pair, becomes broader towards the lateral 

 rectrices, where it is about 12 mm. wide. Total length of adult male (type) about 

 1.51 mm.; wing 69; tail 80; bill 8; tarsus 19. 



Rh. saipanensis resembles very closely Rli. rubrqfrontatd Rams, from the 

 far distant island of Guadalcanar, in the Solomon group of islands, from which it 

 chiefly differs in the greater e.xtent of rnfous on the base of the tail and the slightly 

 wider rufous forehead. 



Oustalet, in his article on what he calls Rh. versicolor, further expresses his 

 opinion that his own Rh. dsirolabi, which he described from Vaiiikoro in the Santa 

 Cruz Archipelago, is the same as Rh. versicolor, thinking it may be its summer 

 dress. I see no reason to accept that view, as Oustalet's theory about the Rhipidurae 

 of Saipan broke down, not to speak of the peculiar distribution at all. Thus I 

 think, without doubt, that Oustalet's Rh. versicolor, as he limited it under No. 7 

 of his article, consi-sts of — 



1. R. cersicolor Hartl. & Finsch, Yap. 



2. R. uraniae Oustalet, Guam. 



3. R. .mipanensis Hartert, Saipan. 



4. R. astrolabi Oustalet, Vanikoro. 



9. Myiagra freycineti Oust. (Oust. I. ji. 194). 



Oust., Bull. Soc. Philom. 18H1 (7), V. p. 7.3; id., Nuturalisle 18811. p. 'iOO ; Wiglesw., Arts 

 Pulynes. p. 24. 



S ad. Maxilla black, mandibula plumbeous. Top of the licad deep metallic 

 blue, with a somewhat slaty hue : rest of uiiper surface dark grey, with a ste<'l-l)lue 

 gloss : winss and tail blackisli, outer webs narrowlv edged with grev, inner webs 



