the Tubinares in the North Atlantic Islands. 457 



drawn up, Avill show that it is very difficult to determine 

 exactly when the breeding- season commences. Padre 

 Schmitz, whose excellent bird diaries cover a period of over 

 seventeen consecutive years (1893-1910) for these islands 

 alone, has furnished us with sufficient data to attack the 

 problem with some chance of arriving at a successful con- 

 clusion. 



The island of Porto Santo is unquestionably the one most 

 favoured by this Petrel upon which to breed, and to this 

 island the majority of Padre Schmitz^s notes refer. 



Either young-in-down or the eggs have been actually 

 taken in the Madeira Group in each of the following 

 months: — January, February, March, April, June, August, 

 September, October, November, and December ; while it will 

 be seen, by glancing at the dates in the above-mentioned 

 schedule, that either incubation of eggs or the rearing of 

 the young takes place in this group of islands in every 

 month of the year save, perhaps, May. 



At first sight it would appear, therefore, that 0. castro 

 has no fixed time in which to breed in this group, but a 

 closer study of the facts has induced me to believe that the 

 bird has two main breeding -seasons. I do not necessarily 

 mean by this that the same birds breed twice in the year, 

 although there is no apparent reason why they should not. 

 It will be seen that June is the first month in the year when 

 nesting may be said to have become general ; the next 

 month in which any quantity oi fresh eggs appear to have 

 been collected is October ; in November the eggs found were 

 mostly incubated ; in December the young are in various 

 stages of development while late birds still have eggs, which 

 accounts for young-in-down being sometimes found as early 

 in the year as February and March. 



We then have two distinct seasons ; the first commencing 

 in June and extending through July, August and September, 

 the second commencing in October and extending through 

 November and December — the late birds of the first season 

 "overlapping^' the early ones of the second. 



In 1899, Padre Schmitz advanced the theorv that the 



