Rouse, Respiration in Pigeons. 503 



sis for deciding as to color-preference was the systematic record. 

 This is given in Table I. Some of the reactions are contra- 

 dictory, but the general result certainly shows a considerable 

 preference for the colors at the blue end of the spectum.^ 



V. Detailed Study of Respiratory Reactions to Light. 



Two distinct classes of respiratory reactions were studied : 

 first, a series of reactions to light of different qualities but of 

 presumably the same intensity ; and, secondly, a series of reac- 

 tions to light of different intensities but of the same quality. 

 The first set of trials was given with a view to determining the 

 relative stimulating value of light of different qualities ; the sec- 

 ond set, to investigate the relation between the reaction and 

 the quantity of the light employed, the stimuli being so 

 applied that the influence of repetition could be observed. Each 

 stimulus was always given as nearly as possible between two 

 respirations, or at the beginning of the inspiration, and was 

 continued two seconds. 



Various modifications of the breathing curve were noted, but 

 special attention was paid to alterations in quickness.' The 

 rate per minute was estimated for each of the five respirations 

 preceding the beginning of the stimulus, and for each of the 

 ten following it, and the average rate of the two respirations 

 immediately prior to the stimulus was used as a "normal rate," 

 or basis, upon which to calculate the amount of rate alteration. 

 It is absolutely necessary to estimate an animal's breathing rate 

 immediately before each test, since the rapidity of normal res- 

 piration is liable to slight variations, not only from day to day, 

 but from one part of an experiment to another. Some students 

 of human respiration overlook the extreme significance of the 

 normal rate in determining reactions, and compare directly the 



'Graber's negative results [Grundlinien zur Erforschwig des Helligkeits toid 

 Farbensimies der Ticre, 102. Prag, 1884) were probably due to the testing of many 

 animals at once or to failure to change the glasses regularly. 



^When shallowing accompanies quickening, the amount of respiratory activ- 

 ity may be no greater, or even less, than in the previous condition. But since 

 there was generally no depth alteration after the first day, the quickening report- 

 ed may be regarded as a fair measure of the influence of the stimulus. 



