338 THE AMERICAN NATURALIST [ Vol. XLII 



weight, and by other means. It seemed worth while to 

 ascertain what effects, if any, such changes of water 

 density have upon the blood. The results of a few 

 typical experiments are here stated. Hemacytometer 

 determinations were made upon fishes subjected to dif- 

 ferent conditions. The mean blood count of a number of 

 specimens of Fund id us licteruclitus in natural sea water 

 was found to be 2,700,000 red corpuscles per cubic milli- 

 meter. Sea water contains about 3.5 per cent., by weight, 

 of salts in solution. In an experiment in which the salt 

 content of the water was increased to over 6 per cent, by 

 the addition of sea-salt, the proportion of corpuscles had 

 increased to about 3,500,000 per cubic millimeter. In the 

 case of another lot of fishes which were placed in 1,500 

 cc. of sea water, to which 30 grams of sea-salt had been 

 added, it was found after six hours that the number of 

 corpuscles was 2,800,000 per cubic millimeter. In yet 

 another lot which were placed in 1,500 cc. of sea-water 

 plus 60 grams of sea-salt, the blood count, at the end of 

 six hours, gave 3,000,000 corpuscles. In a case in which 

 120 grams of sea-salt were added to 1,500 cc. of sea- 

 water the blood count gave 3,900,000 per cubic millimeter. 

 A number of fishes were placed in sea water artificially 

 strengthened to the point of containing about 9.5 per cent, 

 of salts. After one hour it was found that the fishes 

 had lost in weight, while the blood counts of two speci- 

 mens averaged 2,800,000; at the end of two hours, a 

 further loss in weight was noted in the remaining fishes, 

 and in two specimens the average number of corpuscles 

 counted, was now 3,434,000 per cubic millimeter. Fishes 

 which were placed in distilled water and subjected to 

 the hemacytometer test an hour and a half later yielded 

 about 2,000,000 corpuscles to the cubic millimeter. In 

 two exp'eriments with distilled water there was at first a 

 decrease in the number of corpuscles counted, then a 

 gradual increase up to the normal, which increase later 

 passed above the normal. The question as to whether or 

 not the blood is diluted by the osmotic influx of water 



