June 9, 1887] 



NATURE 



129 



contagium in the milk. This, I am glad to say, is very 

 easily carried out. Heating milk up to 85° C. or 185° 

 F., that is, considerably under the boiling-point, is 

 perfectly sufficient to completely destroy the vitality of 

 the microbe of scarlet fever. In harmony with these ex- 

 periments on the influence of heat on the microbe of 

 scarlet fever, I can quote, besides the observation given 

 above by Dr. Robertson, also the following observations 

 recorded by Dr. Jacob, Medical Officer of Health of High 

 Ashurst and Headley, and reported in 1878, to this effect. 

 Between June i and 7, there were fifteen cases of scarlet 

 fever in three distant houses, the inmates of which had 

 had no communication with infected persons, but had all 

 been supplied with milk from a farm where a certain 

 cowman worked. This cowman had in his family several 

 children ill with scarlet fever. The cowman continued 

 milking the cows during the illness of his children, though 

 he did not himself have the fever, and the milk was not 

 taken into his cottage ; but the point which I wish to bring 

 out is this, that other houses besides those in which scarlet 

 fever had broken out had been supplied with the same 

 milk, but no scarlet fever occurred in them, and why .'' 

 because all these had consumed only the scalded milk. 



I should therefore strongly urge that all milk should 

 be boiled, or at any rate heated to at least 85° C. 

 (that is 185° F.) before being consumed. Judging by 

 the large number of cases of scarlet fever recorded in 

 these milk epidemics, one is justified in saying that a con- 

 siderable percentage of the total number of cases of scarlet 

 fever would have been avoided thereby. Not all, because 

 unfortunately the rules of isolation of patients suffering 

 from scarlet fever are not always rigorously carried out, 

 and therefore infection from person to person will occur. 

 Nor would prevention of scarlatina by milk exclude 

 scarlatina by cream, — cream cannot be easily subjected to 

 heat ; and in the epidemic of scarlet fever that occurred 

 in South Kensington in 1875, and that was investigated by 

 Dr. Buchanan, cream was the vehicle of the contagium. 

 But considering the prominent position that milk occupies 

 in every household with children, the possibility of infec- 

 tion with scarlet fever by raw milk deserves careful 

 attention. 



THE SECOHMMETER. 



A CIRCUIT containing self-induction acts as if it had 

 -^~^ a larger resistance than its true one when a current 

 is started in it, and a smaller resistance when the current 

 is stopped. Hence, if balance be obtained with a Wheat- 

 stone's bridge in the ordinary way, the fact of any of the 

 arms possessing self-induction, or of any one of the arms 

 having a condenser attached to it, will produce no effect 

 on the balance if the battery circuit be rapidly made and 

 broken, provided that the rapidity of make and break be 

 not too great for the currents in the arms of the bridge to 

 reach their steady values each time that the battery circuit 

 is made, and to die away each time that it is broken. If 

 the currents have not time to reach their steady value 

 when the battery circuit is closed, and to die away when it 

 is broken, then self-induction in any one of the arms will 

 produce a disturbance in the balance ; but such a method 

 of measuring a coefficient of self-induction would lead to 

 very complicated formulae, and is not worth developing 

 with the view of obtaining a simple method of measuring 

 self-induction. 



It therefore occurred to us to consider whether, without 

 employing such rapid makes and breaks as would prevent 

 the currents reaching their steady values, the self-induc- 

 tion of a circuit might not be made to act as an apparent 

 steady definite increase of the resistance of that circuit 

 which could be measured in the ordinary way with a 

 Wheatstone's bridge or differential galvanometer ; and 

 by this means the measurement of a coefficient of self- 



induction would simply resolve itself into the measurement 

 of a resistance. And this problem we solved in the 

 following way, in the spring of 1886 : — 



The coil, the coefficient of self-induction of which it is 

 desired to measure, is placed in one of the arms of a 

 Wheatstone's bridge, the three other arms consisting of 

 ordinary doubly-wound resistance coils possessing no 

 appreciable self-induction, and not only is the battery 

 circuit rapidly made and broken, but, in addition, after 

 each closing of the battery circuit the galvanometer circuit 

 of the bridge is either short-circuited or broken, so as to 

 cut out the galvanometer, and after each breaking of the 

 battery circuit the galvanometer circuit is either unshort- 

 circuited or closed again, so that the galvanometer is now 

 operative again. In this way all the successive impulses 

 of the galvanometer needle that are produced on starting 

 the current in the coil with self-induction produce theii 

 cumulative effects, but the successive impulses of the 

 needle that, under ordinary circumstances, would be pro- 

 duced on the needle in the opposite direction are cut out. 

 Hence the self-induction possessed by one of the arms 

 causes that arm to apparently increase in resistance by 

 a definite amount depending on the coefficient of self- 



B 



Fig. I.— Preliminary Apparatus. 



induction and the number of operations performed per 

 minute. This apparent increase of resistance pi'oduces a 

 deflection of the galvanometer which can be noted, and 

 its value ascertained by comparing it with the deflection 

 produced with steady currents when one or more of the 

 arms of the bridge is altered by a known amount, as in 

 making the Rayleigh test. But since the necessity of 

 having to read the deflection limits the speed of perform- 

 ing the double make and break operation, in order that 

 the spot of light may not be sent off the scale, we soon 

 replaced this comparative deflection cumulative method 

 by a much more sensitive zero cumulative method ; and 

 instead of reading the galvanometer deflection we re- 

 establish the balance, and bring the needle back to zero, 

 by altering one or more of the arms of the bridge, as in 

 making an ordinary resistance test with a Wheatstone's 

 bridge. 



The first apparatus for enabling measurements of self- 

 induction to be made in this way was constructed in the 

 spring of 1886, under the superintendence of one of our 

 assistants, Mr. Mather. It consisted of a double com- 

 mutator, shown in Fig. i, the spindle, s, to which the 



