lUO Bl'LLETIN OF THE BUREAU OF FISHERIES. 



Experiment 12 1. 



Catfish {Aine'mrnx n,'hulosns) (mitVd IT (l;iys), weight li'l.tl gninis. Suh 

 water through g-ilLs, fresh on body. After -I'l hours Hsh dead; weiglit 115.1 

 grams. Loss of 5.8 per cent. 

 Evperinvrnt 122. 



Catfi.sh (unfed 17 days), weight 159.8 grains. Fresh water through gills, 

 salt on body. After 2.3^ hours fish dead; weight 167.. 3 grains. (Tainof4.7 

 per eent. 

 This gain is pr<>l)ahly due to the absorption of water after deatii. as disi-u--.(M| 

 earlier. It may be due, howe\er, to the aniiuars ha\ing swallowed water. 



So far none of these experiments have given results actually eontradictory to 

 my \ iew regarding the part played by the gills. The following pair of experiments, 

 however, are not in harmony with the others: 

 K.rj>,;-u,ient 123. 



Catfish (fed 24 hours previously), weight 15(J.9 grams. Salt water through 

 gills, fresh on body (fish inverted). After 1-t Jiours fish still stirring; aftei 

 22 hours fish dead; weight 157.1 grams. Weight practically nnchanged. 

 Experiment 12Jt-. 



("atfish (fed 2-1 hours previously), weight 227 grams. Fi-esh water through 



gills, salt on body (fish inverted). After U hours fish dead: weight 22n..-. 



grams. Loss of nearly 3 per cent. 



Occasional inconsistencies in the results such as these might natui-ally be expected 



under the conditions of experiment. The swallowing of water at any time would 



result in an increa.se in weight, the discharge of feces in a decrease. In this con 



nection it will be noted that the fishes used in the last two experiments had been fed 



2-4 hours 1)efore. 



It is possible, however, that naked-skinned fishes react somewhat difiV^rently from 

 scaly ones, as was sup[)osed by H(>rt. Experiments with such fishes should of course 

 be repeated. . 



If water or .salts Ije taken in or pa.ssed out of the body through the membrane- 

 of the gills, it seems to follow that the blood must be the medium of such exchanges. 

 Such difierences in salinity as were indicated for the fishes used in the analyses above 

 could not, however, have d(>pended entirely upon differences in the salinity of the 

 blood, since the volume of the latter is not sufficient to account for them. Indirectly, 

 then, some or all of the other tissues of the body mu.st have been affected. VVhei-e. 

 in the preceding pages, I have referred to the "bod}' ffuids," I have meant to incluilr. 

 not merely blood and lymph, but the liquid content of the ti.ssues in general. 



It has been assumed by Fredericcj (1885) that in certain invertebrates the osmotic 

 ecjuilibrium which is maintained between the body fluids and the surrounding water 

 is due to the permeability of the gills. Concerning fishes, however, he declares " the 

 gills, so permeable to the gaseous exchanges of respiration, seem on the contrary to 

 con.stitute an almost impassable barrier to the salts dissolved in sea water." 



aOf course living fishos only iiit- liore intended. 



