504 Profs. Ayrton and Perry on the 



in (10), (11), and (12) ; fx being the permeability of the 

 armature iron when the induction is /3/. 



In the above investigation, the iron resistance is supposed 

 to be given, and also the exciting power in ampere-turns. 

 At present this seems a sufficiently practical basis for the 

 calculation, as we usually fix first the size of armature, then 

 arrange as short a magnetic circuit as possible, which must 

 leave sufficient room for the exciting coils. We see, however, 

 that a larger generalization is possible when we know with 

 certainty what is the limiting thickness of winding on the 

 field-magnets. We were allowed to assume in the armature 

 winding that tu^ is constant, a Y being the greatest permanent 

 current per square centimetre allowable. When the winding- 

 is thin, so that the temperature is nearly uniform in the 

 winding, and only when this is the case, is such a rule allow- 

 able, and it is not allowable in cases where the winding is of 

 the thickness usual in field-magnet coils. We are at present 

 experimenting on this subject, but there are considerable 

 difficulties in the way of obtaining practical rules. We have 

 no doubt, however, that there is such a rule as 

 8/A 2 / =p\. 



That is, the greatest number of ampere-turns wdiich can use- 

 fully and permanently be applied on a field-magnet coil of 

 length \ is proportional to X. If /is given, the configuration 

 of the machine enables \ and therefore S 2 / A 2 / to be fixed. 



Then S/A/-T- — =/3 t enables (3^ and /n l to be calculated by 



trial if there is sufficient information about the character of the 



It . 2d I 



iron. — is now made equal to — + - — ,so that the important 

 a 2 a 2 a 1 /x 1 



dimensions are fixed. Should ^ be great, considerations of 

 cost of the field-magnet winding may come in, to alter com- 

 pletely the design of the machine, but for a given configura- 

 tion of machine this is the practical method of working. 



From experiments, the results of which were published 

 before this Society on March 12th, 1887, we came to the 

 conclusion that there is a definite resistance at a joint in a 

 magnetic circuit ; and Professor Ewing found that by cutting 

 a bar into two, four, and eight pieces, the magnetic per- 

 meability seemed to alter from 1220 to 980, 640, and 480 

 respectively, the joints being tooled up in the usual way. 

 When the joints are carefully scraped he also found that 

 they materially increased the magnetic resistance unless con- 

 siderable pressure was applied. Now as a stress of less than 

 one quarter of the ordinary stresses to which wrought iron is 



