EFFECTS UPON FISHES OF CHANGES IN SALINITY OF WATER. 99 



Experiment 115. 



One carp (unfed for 22 days), weight 353.5 grams. Salt water through 

 gills, fresh on body. After 15 hours tish dead; weight 338. 7 grams. Loss of 

 about 4 per cent. 

 Experiment 116. 



One carp (unfed for 22 days), weight 3-1-.2 grams. Fresh water through 

 gills, salt on body. After 14^ hours lish in good health; weight 34.5 grams. 

 Weight practicalh' unchanged. 

 Experiment 117. 



One carp (same specimen as used in llti), weight 33. »! grams. Salt water 

 through gills, fresh on body. After IBA^ hours tish dead; weight 31.1 grams. 

 Loss of 7.4 per cent. 

 Experiment 118. 



One carp (fed preceding day), weight 476.7 grams. Fresh water through 



gills, salt on body. After 7 hours weight 476.8 grams. \\'oight practically 



unchanged. 



It will be noted that a considerable loss of weight occurred in all of those cases 



in which salt water passed through the gills and fresh water over the body, while the 



weight remained practically stationary in those cases in which the conditions were 



rev^ersed. Under such harsh treatment the death of any of these fishes would not 



have occasioned surprise. As a matter of fact, however, the three whose gills were 



bathed by fresh water all remained well till the end of the experiments, while of those 



whose gills were bathed by salt water, two died and one sickened. 



The part played by the gills in the death of fresh-water fishes in salt water has 

 been discussed by Bert (1871, see p. 101). Bert states that "a tench suspended in a 

 vessel full of sea water, the head remaining outside, lives for a long time, if care be 

 taken to bathe the gills with fresh water.'' No weight determinations are recorded. 

 In the case of (practically) naked-skinned fishes, such as the eel and loach, Bert held 

 tliat the general integument made possible the extraction of water from the tissues 

 when the fishes were transferred to salt water. 



Two scaleless species were tested by me according to the method described above. 

 The results were here far less satisfactory than those for the carp. 

 Experiment 119. 



Sea raven (unfed for 8 days), weight 334.8 grams. Salt water through 

 gills, fresh on body. After 6 hours tish dead (for some time); weight 332 

 grams. Loss of about 0.8 per cent. 

 Experiment 120. 



Sea raven (unfed for 7 days), weight 456.3 grams. Fresh water through 



gills, salt on body. After 19^^ hours lish dead (how long?); weight 462.5 



grams. Gain of 1.4 per cent. 



The loss in the former experiment is scarcely significant, being perhaps fully 



accounted for by the removal of nmcus in drying the fish for weighing. The gain in 



the second experiment is what would be expected on hypothesis, though the fact that 



the fish had been dead for an indefinite period lessens greatl}' the force of this 



evidence. 



