498 JOURNAL OF* THE WASHINGTON ACADEMY OF SCIENCES VOL. 11, NO. 20 



the switch is operated by an electro-magnet as soon as a contact button on 

 the boom of the galvanometer reaches one or the other of two fixed contact 

 buttons. As the force with which the galvanometer presses the contact 

 buttons together is very small, it is necessary to employ only a very small cur- 

 rent through its contacts, in order to prevent their sticking. For this purpose 

 a high resistance polarized relay is placed between the galvanometer contacts 

 and the magnet that operates the main switch. As a further precaution, 

 the secondary of a small transformer is connected in series with the coil of the 

 galvanometer, while the primary is connected to the heating circuit in such a 

 way that the operation of the main switch causes a ballistic deflection of the 

 galvanometer towards its central position. All the appliances used are of 

 standard make with a few simple changes. 



The adjustment of the rate at which energy is supplied to the furnace is 

 effected by varying the ratio of the time it is supplied at a higher rate to the 

 time it is supplied at a lower rate. Thus the present type of galvanometer 

 is able to secure the proper ratio in a single cycle while the type employed by 

 White and Adams could, in general, only approximate it after several cycles 

 by a fraction whose numerator and denominator were integers. Under 

 certain conditions this caused a rather large, slow oscillation of the temperature 

 of the furnace. 



In order to vary the temperature of the furnace slowly for taking heating 

 or cooling curves, one of the arms of the bridge is shunted by a variable rheo- 

 stat whose resistance is changed in equal steps at intervals of 30 or 60 seconds. 

 This arrangement has been found to give a steady change of temperature, 

 which, however, is not quite linear. 



The regulator, when used with the ordinary type of platinum resistance 

 furnace, has been found to hold the temperatures up to 1250° constant within 

 0.2° C. for several hours, in spite of variations of 5° in the room temperattue 

 and of 6 per cent in the line voltage. At higher temperatures the tempera- 

 ture falls slowly, but may be maintained within 0.2 ° by an occasional manual 

 adjustment. The temperature coefficient of nichrome or chromel wire is too 

 low for the regulator; but nickel wire may be used up to perhaps 500° and 

 some other alloy, such as alumel, may be found to work at still higher temper- 

 atures. As is the case with the White-Adams regulator, this apparatus, in 

 its present form, is not suitable for use with alternating current or with the 

 pulsating current furnished by the mercury arc rectifier. 



The second paper, on Aerial navigation, was presented by Mr. H. N. Eaton, 

 and was illustrated. It was discussed by Messrs. L. J. Briggs, LittlehalES, 

 A. F. Beal and Pawling. 



Aerial navigation is a recent development as it is only since aircraft 

 have become capable of flying for long distances that the necessity for 

 navigating them as ocean-going vessels are navigated has arisen. Aerial 

 navigation resembles marine navigation in general principles but is more 

 difficult in application owing to the higher speeds of the craft and of the sup- 

 porting medium, to the impossibility of charting the winds as the ocean cur- 

 rents are charted because of the variability of the former, and to the fact 

 that freedom of motion in three dimensions introduces additional problems. 

 Fortunately it is not necessary to determine position as accurately in the 

 case of air-craft as in the case of ships on the ocean, since an error of from 

 10 to 20 miles involves only a few minutes additional flying time. 



There are three general scientific methods of aerial navigation; dead reckon- 



