48 INSTRUCTIONS FOR MAKING PILOT BALLOON OBSERVATIONS 



most satisfactory arrangement is to have the two stations connected 

 by telephone or radio. A timing buzzer or other device (preferably 

 operated b}'^ an electric clock) can easily be installed in the telephone 

 line or radio-transmitter circuit which will serve to mark the release 

 of the balloon and the reading signals for each of the successive min- 

 utes and make possible the simultaneous readings of both instruments. 



153. In case the stations are not connected by telephone or radio, 

 some visual method of signaling should be adopted. Timing the 

 readings by watches is usually found to be unsatisfactory. 



154. Reading angles. — The angles should be read very carefully and 

 estimated to the nearest hundredth of a degree, especially the azimuth 

 angles. It is important that the azimuth angles be read closely, 

 especially when the "C" angle is small. 



155. Recording the data. — The data at the primary station will be 

 entered on left-hand half of Form No. 1110-Aer., table 1, and, when 

 that half of the sheet is filled up, a second sheet will be used, as the 

 second half of the sheet is reserved for the entry of the data of the 

 other station. Thus we have all the observed data for any one minute 

 at both stations, on the same sheet and in the same lines. The data 

 at the secondary station will be entered on the right-hand side of the 

 sheet only. At the completion of the observation the data from the 

 one station will be copied oi> Form No. 1110-Aer. of the other. Also 

 the C angle (difference between azimuth angles) for each minute should 

 be found and copied in the velocity column on the right-hand half of 

 this form, as it is used in the computation. 



156. The balloon will be followed as long as it can be kept in sight. 

 Never should it be abandoned at either station before disappearance, 

 without strong reasons for doing so. However, as soon as the balloon 

 is lost at either station, that fact should be signaled to the other 

 station. In cases where the balloon is lost sight of at one station for 

 an appreciably longer period than at the other, the remainder of the 

 flight beyond the time of disappearance at the one station may be 

 computed by the single- theodolite method, using for the ascensional 

 rate the actual rate of ascent as computed from the two-theodolite 

 portion of the observation. 



157. Following the disappearance of the balloon, and before the 

 theodolite is disturbed from its setting, a check of the levels and 

 orientation will be made the same as in single-theodolite observations. 

 (See par. 93.) Note that the azimuth bearing is read from the same 

 vernier by which the theodolite was oriented. If there are no cor- 

 rections to be made, the "check" will follow in the second line after 

 the last line of observed data. The data will then be plotted and 

 reduced in the same way as in the making of a single-theodolite obser- 

 vation. The methods of plotting vary to some extent, however, and 

 will be taken up in regular order in subsequent paragraphs. 



158. Two-theodolite computation. — The computation of two-theodo- 

 lite observations is tedious and laborious at best and full advantage 

 should be taken, therefore, of all devices and methods which will 

 simplify and reduce the work thereof. With this in mind, the follow- 

 ing methods have been developed and adopted by the Weather 

 Bureau. 



159. Slide rule. — The computation is done with an ordinary 20- 

 inch slide rule of which the sine and tangent scales were specially 

 designed and arranged for the work. The scales are arranged as 



