fm Voltaic Electricity, ^c. 125 



notice a slight falling away from the theoretical attraction," 

 owing no doubt to the gradual approach to the limit of magne- 

 tizability in the small bar of iron. And concerning Exp. IV., it 

 is said that " Here again we have evidences of an approach 

 towards the limit of ma gnetiz ability, for the attractions with a 

 current of 4 are only ten times instead of sixteen times as great as 

 those observed with a current of 1." Experiments of this kind 

 are most difficult to perform in a satisfactory manner ; and cer- 

 tainly these can add little to our knowledge on this point after 

 the appearance of Dub's paper on the subject in Poggendorff 's 

 Annalen*, which furnishes details of very numei'ous and most 

 carefully conducted experiments with various cylindrical electro- 

 magnets and keepers. Much depends on the size and form of 

 the keeper; for Dub found that, by merely changing the form of 

 a keeper (the mass remaining constant), the lifting power of the 

 same electro-magnet varied between limits of 1 and 10 at least. 

 Again, much must depend on the soft iron of the electro-magnet 

 itself; and therefore no great confidence can be placed in the 

 result of Exp. V., which assigns the maximum attractive power 

 pe?- square inch of surface of an electro-magnet. Mr. Joule does 

 not appear to have tried a sufficient variety of forms, both of 

 magnet and keeper, to warrant him in fixing this limit. 



From Experiments I. and II. we see no indication of an ap- 

 proach to the limits of magnetizability ; but with the same cur- 

 rents and magnet we find that this is the reason given for the 

 wide departure of experimental from theoretical attractions. 

 Now from the first two experiments it is manifest that thei'e 

 could only be an approach to the limit of magnetizability in the 

 keeper in Exp. IV. ; consequently a more massive keeper should 

 have been tried afterwards ; and this was the more necessary, as 

 some recent experimenters have denied the existence of such a 

 limit. 



The rule for comparing the lifting powers of two similar elec- 

 tro-magnets does not appear to be by any means satisfactory. 

 As we are not told in what this similarity is supposed to consist, 

 it is impossible to test the theoretical correctness of the rule. 

 The two electro-magnets compared by Mr. Joule wore provided 

 with 60 and 100 lbs. of coils of copper wire, and with 10 and 16 

 cells rcsjjcctively. Therefore by the rule, the attractive powers 

 ought to be as 60 X 10 : 100 X 16 = 6 : 16 = 1 : 267. The attrac- 

 tive ])owcrs were found to be in the following ratios : — 

 At a distance of \ inch as 180 : 976 = 1 : 2-03 

 I ... 168:320 = 1:1-90 

 1 ... 77:140 = 1:1-82 

 St. John's College, Cambridge. 



* Phil. Mag., March 1851. 



