518 The Rev. T. R. Robinson's Experimental Researches on the 



decreasing below this in accordance with my former observations, which led 

 me to conclude that this occurs for a certain ^ which I call B, possessing some 

 notable properties.* The residual magnetism, or, as M. Joule terms it, the 

 " Set of the Magnet," is always what is left by a single excitement ; for in each 

 of these experiments the polarity was reversed : its amount in respect of the 

 total L is insignificant. The force when the brass plates are interposed is of 

 considerable amount with this magnet for its higher powers, and follows a very 

 different law from that which prevails in actual contact, being as ■^-. This, as is 

 obvious, implies that the magnetism excited when the poles are not connected is sim- 

 ply as -f, — afact which apparently depends on its comparatively small intensity. 

 As I wished to ascertain the properties of an electro-magnet composed alto- 

 gether, or in part, of hard steel, one was prepared, and another exactly similar was 

 made of iron to serve as a point of comparison. As I doubted the possibility of 

 hardening viniformly masses of steel as large as those of either of the magnets 

 already described, these were made of smaller dimensions, the cylinders 1' dia- 

 meter and 7'-85 long, solid except that they have at the top holes 0-25 diameter 

 and 0-5 deep, to receive thermometers. The base and keeper are 1' high by l'-5 

 broad, and the distance of the centres is 6'. As the steel cylinders fit the base and 



tatlon. Comparing the power of the hollow S^-inch magnet (ii. p. 512) with one of 2' solid, and 

 the same length (given ii. p. 520, No. 62), I find — 



Observed powers as, ... . 10'65 : 4. 



Theoretic, 10'43 : 4. 



Section, 10-00 : 4. 



The hollow 2-inch gives, compared to a solid one of the same diameter and length (tp. 295, Note), 



Observed powers as, ... . 3'03 : 4. 



Theoretic, 318 : 4. 



Section, 3-00 : 4. 



And the solid inch compared with the above No. 62 — 



Observed powers as, ... . 1 09 : 4. 



Theoretic 1-22 : 4. 



Section, 1-00 : 4. 



As in the hollow magnets the iron is only 0''5 thick, we may suppose it is all uniformly mag- 

 netized ; and the agreement shows that the other must be so too. 



* When the keeper is separated from the poles by discs of paper 0''0041 thick, A = 111 '44 

 with yfr = 900. With the interval 0-12 there is no sensible value of it. 



