78 SEA-BIRDS 



But on 20 February 1906, two hundred and seventy-seven years after 

 the existence of the pecuUar bird had been last mentioned, a black- 

 capped petrel with white under-parts was discovered in a rock- 

 crevice on Castle Island, Bermuda, by Louis L. Mowbray. It was 

 like no other petrel that had ever been seen, though it bore some 

 resemblance to Pterodroma hasitata. With J. T. Nichols, Mowbray 

 described the new bird in 191 6 as Pterodroma cahow. In the same year, 

 though a few months later, R. W. Shufeldt described a new petrel 

 from a large collection of bones gathered from the floors of the old 

 bird-caves in Bermuda by Mowbray and others; he also believed 

 that this was the old, real cahow. The issue had by then become 

 somewhat confused, for Bermuda then harboured breeding colonies 

 of shearwaters. There is no doubt that some of the nineteenth-century 

 accounts of 'cahows' on Bermuda — such as those of J. M.Jones (1859) 

 and S. G. Reid (1884) — refer to Audubon's shearwater, Puffinus Vherm- 

 inieri. Both Jones and Reid alluded to their birds, not inappropriately, 

 as Puffinus obscurus, an old synonym of both P. Vherminieri and P. 

 assimilis. 



On 8 June 1935 a boy on a bicycle brought a bird to William 

 Beebe (1935) in his research laboratory at New Nonsuch, Bermuda, 

 from the lighthouse-keeper at St. David's. Beebe sent it to R. C. 

 Murphy, who confirmed that it was the second known specimen 

 o^ Pterodroma cahow. It was a young bird, probably only a few days out 

 of the burrow, and had died by flying against the light. The bones 

 of this bird were, rather luckily, preserved; and were identical with 

 the sub-fossil and recent cave-floor material described by Shufeldt. 

 So there is no doubt that Pterodroma cahow (frontispiece) is the old 

 cahow; and that it survived and bred in 1935. 



In June 1941 a cahow was killed by striking a telephone- wire on 

 Bermuda. In March 1945 F. T. Hall, stationed on Bermuda, found 

 old bones, fragments of birds apparently killed by rats, and a partly 

 disintegrated adult that had floated ashore. 



Murphy and Mowbray (1951) who recount this, were responsible 

 for the uncovering of the cahow's present breeding-grounds in 1 95 1 . 

 Study of this group of petrels had suggested that the cahow might 

 have an early breeding-season, and their exploration of the islands 

 off* Castle Roads occupied the period from 25 January to 10 February. 

 On 28 January they found a cahow on an egg at the end of a six-foot 

 horizontal burrow at the rear of a rocky niche on a much-eroded islet 

 Eventually they found three occupied islets, one with perhaps nine. 



