1891.] PLEISTOCENK BIRD-REMAINS. 475 



The largest species of Puffinus is indicated by the right humerus, 

 represented in figs. 13, 13 a of Plate XXXVlt., which has a total 

 length of 0,115. The proximal end of another right humerus 

 indicates a slightly larger Ijird ; and there are likewise a perfect 

 radius and ulna according in size witii the humeri, while there are 

 also examples of the metacarpus. The skull is indicated hy tlie 

 imperfect rostrum, and the hinder part of the left mandibular ramus. 

 The other bones from Tavolara which I refer to this species include 

 the right coracoid and two examples of the femur. There is also 

 a femur probably referable to this species from Monte San Giovanai. 

 The figured humerus is slightly longer than the one of P. cinereus, 

 Stephens, figured by Milne-Edwards in his ' Rech. Oiseaux de la 

 France,' pi. 52. fig. 7 ; and since it agrees very closely with the corre- 

 spoudiug bone of a skeleton oiP.fuWjinosus, 1 am inclined to refer it 

 to that species, which now inhabits South Europe. The tibio-tarsus 

 of that species has a comiiaratively short cnemial crest, like that of 

 the specimen represented in fig. 14 of the Plate. 



The second species of the genus is represented by the right tibio- 

 tarsus and tarso-metatarsus, drawn'in'figs. 14, 14 a of Plate XXXVII., 

 as well as by the rostrum and portions of the mandible, together 

 probably with some bones not easy to distinguish from those of the 

 third species. The tibio-tarsus is very considerably larger than that 

 of P. obscurus (Milne-Edwards, op. cit. pi. 51. figs. 14, 15), so that 

 these specimens could not possibly have belonged to the still larger 

 P.fuliginosus. In the small relative height of the cnemial crest 

 this type accords with P.fuliginosus, and also with that of P. chloro- 

 rliynchus, Less., of Madagascar, &c., the latter being, however, 

 considerably smaller than the specimen before us. The figured 

 tarso-melatarsus accords fairly well in relative size with the tibio- 

 tarsus (although there are some sliglitly smaller specimens of the 

 opposite side agreeing still better in this respect), and calls for no 

 special mention. The species to which these bones belonged may 

 probably be regarded as one allied to P. clilororhjnohus. 



The third species of Puffinus, as typified by the tibio-tarsus and 

 tarso-metatarsus represented in figs 16, 1 6a of Plate XXXVII., is 

 distinguished from the preceding not only by its inferior size, but also 

 by the much greater length of the cnemial crest of the tibio-tarsus. 

 To the same species may be referred the type of humerus repre- 

 sented in figs. 15, 15 a of the Plate, as well as a coracoid and several 

 examples of the radius, ulna, and metacarpus. There are likewise 

 several more or less imperfect specimens of the cranium and man- 

 dible belonging to this species. 



The tibio-tarsus accords in all respects with the corresponding bone 

 of a skeleton of P. anglorum preserved in the Prosector's room at the 

 Society's Gardens ; and I should have no hesitation in referring the 

 fossil form to that species were there not some doubt whether the 

 Mediterranean representative of this Shearwater does not form a 

 distinct species (P. yelkouan, Acerbi ^). There can, however, I think 

 be no doubt but that the fossil belongs to one or other of these two 



forms. 



' See Yai-rell's ' British Birds,' ed. 4, \ ol. iv. pp. 25, 26. 



