December i6, 1920] 



NATURE 



505 



mounted outside the aeroplane, and spun at 

 10,000 to 15,000 revolutions a minute by the 

 relative wind. The central cylindrical portion 

 passes inwards through the side of the aeroplane 

 and carries the dial, which is in front of the pilot. 

 By means of a cam marked / the tension of the 

 control spring e can be varied at will over a large 

 range, and thus the sensitiveness of the instru- 

 ment can be changed to suit different machines 

 or different atmospheric conditions. 



Another gyroscopic instrument which has been 

 designed during and since the war is the gyro- 



VCNToW 



1 — ^ A i. , -. >|[, A, Vcniuri lab« cKpo^«(l to wiuU , i., -......., «a 



' • - r.-. ) sfe-i '>/ wind nilt C ; d, pipi conneninv throat of Ventari to diapbrmgni t 



' r,if. h»-t .,11.1 wh^el. 



scDpic rudder control. This is intended to relievc 

 Ihe strain on the pilot during a long flight by con- 

 trolling the aeroplane's direction. With it the 

 pilot need only check the course once every five 

 to ten minutes. The instrument consists of a 

 1,'yro.scopc hung freely in gimbals and carefully 

 lialancod. With good workmanship such a gyro- 

 scope will keep its direction within a degree in 

 ten minutes. Any movement of the aeroplane 

 relative to the gyroscope immediately operates a 

 valve controlling a pneumatic servomotor, which 

 puts over the rudder so as to turn the aeroplane 



''-68. VOL. 106] 



back to its original c-ourse. It is also fitted with 

 a simple "follow-up gear," so that the deflection 

 of the rudder is proportional to the amount the 

 aeroplane is off its true course. The apparatus is 

 shown diagrammatically in Pig. 2. The gyro- 

 scopes are about 3 in. in diameter, and spun 

 at about 15,000 revolutions a minute by com- 

 pressed air. 



Until recently the only method of finding the 

 distance an aeroplane had travelled through the 

 air was to note the length of time flown and the 

 average reading of the air-speed indicator. The 

 latter reading, being dependent on the 

 density of the air, must be corrected for 

 the density at the height flown, as well iis 

 for a small error due to interference of the 

 air-flow by the neighbouring' parts of the 

 aeroplane. To eliminate this trouble, an 

 air log was designed, which should record 

 continuously the actual distance travelled 

 through the air. This is shown diagram- 

 matically in Fig. 3. A windmill type of 

 anemometer is employed, since this Sas 

 the advantage that it needs no correct iai 

 for change of density. .As the windmill 

 must necessarily be fixed on an exposed 

 part of the aeroplane, it must be made to 

 transmit its indications to the pilot. This 

 is conveniently arranged by mounting a 

 \'enturi tube alongside the windmill. The 

 exit of this Venturi tube is alternately 

 opened and closed by a rotating disc, say 

 once every 1000 revolutions of the wind- 

 mill. The pressure at the throat of the 

 Venturi tube is thus alternately above and 

 below atmospheric pressure, and a pipe led 

 from the Venturi tube to the indicator 

 transmits these alternations, of pressure, 

 which operate the counting gear through 

 the agency of a small diaphragm. The 

 addition of a small adjustable baffle behind 

 the windmill allows the interference error 

 to be eliminated once for all for any aero- 

 plane. 



The u.se of ordinary sextants in the air 

 is largely ruled out by the fact that the 

 true horizon is nearly always obscured by 

 haze. The top of this layer of haze is 

 generally nearly horizontal, and has, there- 

 fore, sometimes been used instead of the 

 real horizon. As, however, it may occa- 

 sionally be inclined to the horizontal by a 

 degree or more, such results are untnistworthy. 



The use of bubble sextants has been attended 

 by greater success than was at first expected. On 

 aerc)planes the mean of six readings will generally 

 give the altitude of the sun or star correctly within 

 about ten or fifteen minutes of arc. On airships 

 much better accuracv seems [jossiWe. The design 

 of bubble sextants for use in the air is largely a 

 matter of making them convenient to u.se. The 

 image of the sun and bubble must move together 

 across the field if the instrument be tilted, and the 

 size of the bubble must be adjustable. • The general 



