SIBLEY: ORNITHOLOGY 639 



to new locations since AVorld War II. The Rossitten group, under Ernst Schiiz, 

 is now at Radolfzell on Lake Constance, while the Helgoland station has moved 

 to Wilhelmshavn and is directed by Rudolf Drost. The first bird observation sta- 

 tion in Sweden was founded at Ottenby in 1945 and others (e.g., Fair Isle, Isle 

 of May, Skokholm) have been intermittently active in the British Isles, 



The information derived from banding includes much in addition to data 

 concerning migratory routes. Knowledge of the dispersal of juveniles, sex ratios, 

 speed of flight during migration, longevit}^, plumage change in relation to age, 

 and diseases and parasites are among the items to which banding has made a 

 direct or indirect contribution. 



Thanks to the banding technique the mysteries of bird migration were fewer 

 in 1920 than they had been in 1900, but at least two major problems remained 

 unsolved. AVhat was the stimulus which started a bird off on its migratory flight 

 with such remarkable precision, and how did the migrating bird find its way? 

 These questions demanded experimental investigation. The precision with which 

 migratory birds arrived at a given point year after year was proof that the 

 timing device which provided the stimulus was equally precise. The annual 

 cycle of weather, seasonal variation in food supply, and other phenomena had 

 been suggested as the source of the stimulus. These w^ere too variable to account 

 for the regularity of migration. When Professor AVilliam Rowan of the Univer- 

 sity of Alberta began his investigation into the problem, he had logically settled 

 upon the annual cycle of changing day-length as the only apparent cyclic phe- 

 nomenon with the necessary degree of precision. This hypothesis he set out to 

 test. In the fall of 1924 Rowan trapped southbound slate-colored juncos pass- 

 ing through Edmonton. The birds were caged in outdoor aviaries, one of which 

 contained an electric light. The experimental procedure was classically simple. 

 Beginning on November 1 the light in the experimental cage was left burning 

 for 71/2 minutes after dark. A daily increment of 71/2 minutes was added until 

 December 3, when the increment was reduced to 5 minutes. By December 15 this 

 procedure resulted in the light remaining on until 11:00 p. m. The increases 

 were then discontinued, the light going out at 11:00 p. m. until January 9, 

 when the experiment was terminated. Although their environment was that of 

 a Canadian winter the gonads of the experimental birds had attained the maxi- 

 mum breeding size, and the males were in full song. The gonads of the control 

 birds in the unlighted cage were at the winter minimum size. Here indeed was 

 proof of the effect of photoperiodism on the sexual cycle. Rowan's little book 

 The Riddle of Migration (1931) summarizes his experiments. Although some 

 of Rowan's conclusions concerning the relationship between the gonad cycle 

 and the migratory impulse have been modified, his experiment started the inten- 

 sive investigations, such as those of Wolf son (1945), which have led to an under- 

 standing of the annual stimulus for migration. The present state of knowledge 

 has been summarized by Earner (1950), who proposes a working hypothesis 

 which attempts to reconcile many seemingly divergent facts and suppositions. 

 This hypothesis, somewhat simplified, states that twice each year migratory spe- 

 cies of birds come into a distinct physiological condition which places the bird 

 in a "disposition to migrate." This is indicated by the deposition of fat in many 

 species. The gonads begin to increase in size and a condition of "restlessness," 

 which is especially noticeable in caged birds, is evident. 



