292 



THE AMERICAN NATURALIST [\''ol. XLI 



I watched the respiration of skates in aquaria supplied with running 

 water, observing the fish at times when they had not been dis- 

 turbed in any way for several hours. At such times the rate of 

 respiration was always slow, — usually from 22 to 30 inspirations 

 per minute. At fairly regular but long intervals there occurred 

 a break in the regular alternation of inspiration and expiration. 

 This break was brought about in the following way. Immediately 

 after an inspiration (and therefore in a period ordinarily marked 

 by a contraction of the pharynx with closed spiracle and open 

 gill clefts) the spiracle remained open and the gill clefts remained 

 tightly closed while a particularly vigorous contraction of the 

 pharynx caused the contained water to be ejected forcibly from 

 the spiracle. It is apparently by muscular action that the gills 

 are kept closed during the spouting, since the pressure of tlie 

 water in the gill chambers would tend to force open the extei nal 

 valves. During the spouting the mouth was open, as it is (huing 

 inspiration, and some water escaped from it, but very little as 

 compared with the amount ejected from the spiracle. It is doubt- 

 less due to the respiratory valve that the outflow from the mouth is 

 not greater. The contraction which caused the spouting was 

 immediately followed by an expansion of the pharynx, thv spiracle 

 still remaining open and the gill clefts closed, and respiration 

 then proceeded in the usual way. In animals which had Ih cii at 

 rest for several hours, the rate of rcs|)iratioii l)cin<^ \\\vu at its 

 lowest, the spouting occurred at iiitcr\al> of five to ten iiiimites. 



Having found that spouting is a feature i)t' normal respiration 

 in a resting fish, I next sought to discover what part the spouting 

 plays in the respiratory process. With this end in view, I observed 

 the fish under other conditions than rest. 



Effrds of Exercise.— The rate of respiration in a fish varies 

 with the degrec^ of activity. To induce rapid respiration 1 caused 

 the fi^h U, take exen i.e. This was etfectively .lone l,y gnisping 

 and 1 lli.th I t fill I I thetaih Th<' iuo>t violent ellofts 



(Ij A skate had been un<listnrl)<-.l over ni.-lit in an acpiariuin 

 supplied with running sea water. When (ir>t observed in the 



