EXPERIMENT STATIONS. 251 



problem l)y the law of correction, b}'^ which wo can correct 

 our results as an astronomer corrects his results, in order to 

 make them agree and bring them down to the deliniteness 

 whieh is retjuired. 



Now, in order to get at these true results, it takes money, 

 takes time, takes skill and talent ; and here I am coming to 

 what I want to say to this Massachusetts audience. In order 

 to make any experiment station a success, 3'ou want to give 

 the men employed there money enough to enable them to 

 study these things in a thorough way, or you cannot expect 

 to learn much from that sttition. If you want your Massa- 

 chusetts Experiment Station to do good work, it must have 

 enough funds given it to remove it from criticism. It is not 

 fair, it is not proper, to criticise the work done at an estab- 

 lishment, where so much is expected from them and where 

 no opportunities are given them tor doing any work. An 

 annual appropriation of $20,000 is none too much. There 

 are subjects coming up every little while, which must bo in- 

 vestigated, and which may require the expenditure of nioncy. 

 If you are working close to the appropriation, you will not 

 be able to do that work. The station should have a full 

 treasury all the time. My own plan is to keep one or two 

 or three thousand dollars leeway ; not to expend the appro- 

 priation within two or three thousand dollars, if I can arrange 

 it, so as to have money in the treasury to meet any unoxpocted 

 demand which may call for additional labor or skill in doing 

 the work. We only spent $17,000 last year of our appro- 

 priation of $20,000, and this year we shall spend about $iO,- 

 000. If we had $17,000 appropriated, I should make my 

 plans to spend but $15,000, so as to have $2,000 leeway. 

 When you give your experiment station the paltry sum of 

 $5,000, ib is hardly enough to overcome friction and pay its 

 employees ; 3^ou have so hampered it that it ouglit not to be 

 the subject of criticism. You ought either to st()[) that ap- 

 propriation and let the station die, or give it somelhing 

 more. There is nothing gained l)y hampering this work by 

 insufficient means. The only (juestion is, whether it is svorth 

 the expense. If you want an engine to move a train of cars, 

 yon cannot do it if you have only steam enough in the boiler 

 to overcome friction, and the whole expense up to this point 



