[11] EXPLORATIONS ON COLUMBIA RIVEil. 247 



but is also a standing temi^tation to the operator to hatch more eggs 

 than it is capable of doing. The result of this, of course, is usually a 

 disastrous loss. Then, again, if a dry season should come, or one 

 otherwise unfavorable to the supply of water, that season might prove 

 a,lmost an entire failure. It is consequently hardly desirable to under- 

 take the hatching of salmon on a large scale without being sure of 

 having plenty of water. Indeed, to be short of water in the hatching 

 season is so annoying, not only from the causes just mentioned, but for 

 various other reasons, that I would not want to have anything to do 

 with a hatchery that did not have a large, superabundant, and unfail- 

 ing supply of water. The first station of the United States Fish Com- 

 mission on the McOloud River was a very good example of the absence 

 of this condition. The station combined almost all the conditions ex- 

 cept this one, and those who carried it on the only season that it was 

 in existence will never forget the great care and anxiety that were 

 caused by the insufficient supply of water, or the alarm that was 

 always felt when hot and dry weather shrunk the little supply that we 

 had, and there was danger that all the eggs would be lost in conse- 

 quence. 



4. Convenience op location for obtaining suitable water. — 

 This condition may seem at first sight to be identical with the last, but 

 a second look will show that it is not only a different one, but one that 

 may often be wanting where the other is unexceptionable. This is not 

 an uncommon occurrence. You may have a large river full of salmon, 

 plenty of water, and plentj' of fish, and not be able to use any of the 

 water for hatching the eggs that are taken. I refer now to an auto- 

 matically provided supply of water. For I am, of course, aware that 

 wherever there is water it can be raised to any reasonable height by 

 steam-pumps and other agencies ; but steam-pumps are expensive to 

 begin with ; they involve a current expenditure in the running of them, 

 and are never wholly free from risk. It is consequently always desir- 

 able to provide the water for a hatching station automatically, and no 

 place where this cannot be done can be said to combine all the condi- 

 tiops desirable for a successful hatching station. 



It sometimes happens that sufficient water can be brought to the 

 hatching house without much inconvenience, but owing to its becoming- 

 warm or muddy on its way it may be unfit to use after it reaches there. 

 It also happens sometimes that, although abundant and suitable water 

 for the hatching house may be very near, there may be great incon- 

 venience and risk in bringing it to where it is wanted. It is conse- 

 quently quite important in selecting a hatching station to find a place 

 where the water can not only be found m abundant supply, but where it 

 can be conveniently brought to the hatching house in suitable condition. 

 Probably more ingenuity has been exercised by fish-breeders in their 

 contrivances for bringing suitable water to their hatching houses than 

 in any other department of their operations. Windmills, steam-pumps, 



