6 



second was of the usual size, and the third was a very small one. As the nest was placed very 

 far from the trunk of the tree, the eggs had to be got out with a sort of landing-net, and the 

 small egg fell and was injured. I have kept these three eggs as curiosities. If the eggs of this 

 Hawk are taken when quite fresh, others are immediately deposited, as is the case with the 

 Common Sparrow-Hawk. If, however, the eggs are much incubated, and the season advanced, 

 as was the case this year, they seldom lay again. This year only two pairs, that I know of, laid a 

 second time, though I am convinced that this Hawk sometimes lays a third time if the second 

 sitting is taken immediately." 



In Dresser's collection are three eggs of this bird, taken by Dr. Kriiper at Olympus on the 

 12th of June 1870. In size they measure lf-§ inch by 1^ inch, and in colour are pale yellowish- 

 white, almost the colour of old ivory, but appear to have contracted the yellow tinge from nest- 

 stains. Compared with eggs of Accipiter badius from North-west India they are easily distin- 

 guishable, as the latter eggs, although about similar in size, are of a pure bluish-white colour, 

 and have a rather rough grain, like a Marsh-Harrier's egg, whereas those of Accipiter brevipes 

 have a rather smooth shiny surface. 



Dr. E. Rey, of Eathmannsdorf, near Halle, who is the fortunate possessor of the first eggs 

 of this Hawk taken on European soil, at Puonica, in Servia, on the 13th of May 1869, writes us 

 that of these and other specimens procured by Dr. Th. Kriiper, in Macedonia, nineteen eggs in 

 all, the average size, in millimetres, is 39 - 8 by 3T4, the largest measuring 43*5 by 32 - 5, and the 

 smallest 37 by 30. 



We have said very little on the subject of the supposed identity of this species with Accipiter 

 badius and A. sphenurus, with both of which it has been erroneously united. Great confusion 

 exists as to the different species of small Sparrow-Hawks of the present group ; but it is, in our 

 opinion, a mistake to refer them all to one species, as Drs. Finsch and Hartlaub have done, or to 

 consider them varieties of one common form, as Dr. Bree has proposed. To unite every variation 

 in form under one head in a great measure renders null and void any attempt at a correct 

 appreciation of the geographical distribution of birds. We have examined, thanks to the kind- 

 ness of many good friends, a fine series of all the species which have been merged by authors 

 under the heading of Accipiter badius, and we believe that there are at least five different species, 

 each possessing a clearly characterized geographical range, as follows : — 



1. Accipiter badius, with the female conspicuously larger than the male. Hab. India. 



2. A. brevipes. Largest ; sexes very little differing in size. Female resembling that of 

 A. nisus. Hab. Central Russia southwards to Syria. 



3. A. sphenurus, with which A. brachydactylus is identical. Very similar to A. badius, but 

 not exhibiting so much difference in the proportions of the sexes : the female never so large as 

 in true A. badius. Hab. Senegal to North-east Africa. 



4. A. polyzonoides. Smallest : never so rufous as the others. Hab. Damaraland to Zambesi. 



5. A. madagascariensis. Hab. Madagascar and the Comoro Islands. 



The adult and young males figured in our first Plate of this species are taken from Mace- 

 donian specimens in our collection received from Dr. Kriiper. The second Plate represents an adult 

 female and a young male in change, the former being in the Norwich Museum, and kindly lent to 

 us by the authorities, the latter killed by Canon Tristram, in Palestine, and still in his collection. 



