20 Magnetical Dip in the United States. 



on the face of the published results and must be apparent by a 

 COM/? d^ccil to every professional observer. Why then such a la- 

 bored table of them ? But they all amount to nothing, literally 

 they amoimt to nothing ; for in my mode of manipulating, adopt- 

 ed since Sept. 1839, they disappear from the " mean results ;" like 

 the zero error they are mostly corrected by the reversals. I could 

 point out what I conceive to be the cause of these vacillations, 

 and the mode in which I have succeeded in merging them ; but 



1 have neither time nor an inclination at present, to go into details 

 on these points, I had intended to publish the result of my ex- 

 perience, but the duty to which I am here called, is very far from 

 that of being an instructor of others. I certainly ought not to be 

 forward in volunteering my services in that capacity. 



The evidence which convinces me of the correctness of the 

 mean results of my observations, is of the most popular kind ; a sort 

 of evidence which every body can understand, equally applicable 

 to moral and to physical snbjects, entirely independent of hypo- 

 thesis and assumptions of any kind, and as ancient as the penta- 

 teuch. It is simply the agreement of two independent physical 

 witnesses to the same result, under repeated and varied trials. 

 The dip is twice taken, at each place, by a mean of 16 readings of 

 two separate needles. Out of 16 pairs of mean results, at differ- 

 ent places, published in the last number of this Journal, there 

 were but two, where the separate needles disagreed more than 1' 

 of a degree. What is common sense to infer from this ? that 

 there is an error of "30 minutes" in any one of them, both the 

 needles taking a fancy, or conspiring, to agree to the same false- 

 hood. Prof. Loomis will not dispute the facts above stated, 

 and we submit the case, whether they do not authorize the con- 

 clusion which I have drawn, " that the results are true to within 



2 or 3 minutes." 



Prof. Loomis has taken no notice of this part of the evidence 

 of the accuracy of the mean results of my observations. What 

 must I infer from this ?* 



Lastly, Prof. Loomis cites a discrepancy between the dip at 

 Cincinnati, as observed by me in 1837, and the same observed 

 more recently, as proof that my observations are erroneous. I 

 had myself published the fact of that discrepancy, and had it un- 



* Should the two needles each have pivots of any other form than that of the cylinder 

 and be alike, and alike placed in reference to the axis of the needle, they might agree in 

 an erroneous indication. But such a concurrence can scarcely happen. 



