LIGHT AND MOVEMENT 



65 



(Figs. 39 and 40). By keeping the birds in conditions wherein tlie sun and 

 sky were excluded for a number of days before release, consistent errors were 

 made which could only be explained on the supposition that the birds were 

 failing to correct for the seasonal variation in the sun's altitude from which they 

 derived their measurement of latitude. By de-synchronizing the day-night 

 rhythm before release by arranging an artificial day beginning and ending a few 

 hours earlier or later than normal, errors in longitude were made which could 

 be explained on the basis of a disturbance of an inherent time-sense based on 



Home 

 -noon L, 



\ry- ^c^i^"®"^* posicion 



' ^- 2t local noon 





o^;:o-^^ 



•Decrease in 

 azimuths time 



Fig. 41. — Diagram Illustrating the Hypothesis of Sun Navigation. 

 Released to north and west of home. See text. (Tlie diagram is not to 

 scale.) (After G. V. T. Matthews.) 



regular light-dark sequences ; they flew in a false direction — too far east after 

 an advanced day, too far west after a retarded day. That' the direction is 

 determined by the incident light was strikingly shown in Kramer's (1952) 

 experiments with migrating starlings : when the light was deflected by 90^ by 

 mirrors, the birds' flight was equally deflected and in the same direction, 



Wilkinson's hypothesis is illustrated in Fig. 41. Briefly, the sun's arc 

 is observed over a small excursion and from this its position at local noon and 

 the geographical south are extrapolated ; the latitude is determined by the 

 difference between the observed noon altitude and the remembered noon 

 altitude at home. The difference in longitude is derived joartly by comparison 

 with the home position in azimuth at local noon combined with an estimation 

 of time in the diurnal night-day cycle. This, although it is not yet experimentally 



S.O. — VOL. I. 



