Farr and Macleod. Artesian Wafer and Ilatc/iiw/ of Trout. 57 



A glance at these figures will show that, while the two top boxes are 

 practically identical in nitrogen-content, that supplying the set of five 

 trays has a little less oxygen (046 c.c. per liti'e), but has an emanation 

 content 1-6 times that of the other. 



Under the headings " Eggs" and '" Blue Swelling" are given the numbers 

 of deaths as eggs, or in the yolk-sac stage due to blue swelling, as the case 

 may be. The figures in brackets indicate the same quantities in check 

 boxes occupying parallel positions to those in the same square. Only a 

 few such parallel boxes were used in the one set of trays. 



The last columns in each set are not strictly comparable. They refer 

 to the same class of fry which were hatched from " pond " and not " wild " 

 ova, and in the longer set of trays we had only a few eggs at the end of the 

 hatching season to experiment with, and the results were raised to the 

 2,500 basis for comparison. The discussion of these figures in conjunction 

 with last year's results show, we think, that both the death of eggs and 

 also blue swelling decrease as the water becomes more aerated, and this 

 seems to be more the case in the set of five boxes than in the set of eight. 

 It may not be considered to be very marked as regards eggs in either set 

 of the brown-trout ova, but it was very evident in the same set of boxes 

 last year, when 7,500 eggs were hatched in each box. Taking the figures as 

 they are, while it cannot be said there is any falling-off in the number of 

 deaths in the eggs in the set of eight boxes, there is a slight falling-ofi" in 

 the set of five boxes, and a distinctly greater mortality on the whole, 

 especially in the boxes nearest the well. 



As regards blue swelling, whereas in both sets of rainbow fry. and in 

 the set of five trays with the 2,000 brown trout, there is a marked falling-oft' 

 from box to box, yet in the set of eight boxes it seems that the mortality, 

 as with the eggs, is practically constant. 



It would appear that, taken as a whole, the figures obtained last year 

 and these indicate that both these troubles are due to want of aeration, 

 though it cannot, of course, be denied that some eggs would die even though 

 river- water perfectly aerated were used. 



The hatching season is short, and the whole of the hatchery cannot, 

 of course, be handed over to experiments of this kind, and hence the results 

 are meagre, and conclusions as to the particular constituent in defect or 

 excess which is the cause of the mortality must be drawn with caution. 



If it be admitted that the mortality depends to some extent on the 

 aeration, then we think it can only be due to defect of oxygen or excess of 

 radium-emanation. Besides the improbability of so inert a gas as nitrogen 

 affecting the fish, the very small percentage differences between the nitrogen- 

 contents of the two top boxes would further reduce the probable effect of 

 nitrogen. 



At present it is practically impossible to decide between a defect of 

 oxygen and an excess of radium-emanation. Both these tend to disappear 

 as the water is aerated, so that the two cannot be well separated. 



We hope we may be able to perform crucial experiments by confining 

 goldfish in a limited sup2)ly of water containing sufficient radium-salts to 

 produce the emanation in quantities comparable with that in the wells. 

 By such means one of the two factors could be eliminated. 



