INSTRUMENTS AND METHODS. 271 



Considerable thought was given in England to devising some method for retrieving the 

 instruments sent up with the balloons. There was little difficulty with the method to be 

 used in the summer, and as this was very successful it will be described first. The details 

 of the method were actually worked out in the Antarctic, but it will be unnecessary to 

 describe the experiments made. There is nothing new in the method as a similar method 

 has been used by Hergesell over the ocean. The instrument was attached to the balloon 

 filled with hydrogen by means of a slow match which was of such a length that it burnt 

 away and detached the instrument after the time required. To the instrument itself a second 

 balloon filled with air was attached and fell with it. This balloon was of red indiambber 

 and was blown up to bs about 4.5 cm. in diameter when at ground-level. The function of 

 this balloon was simply to mark the position of the fallen instrument. On hberating the 

 large balloon with instrument and small red balloon attached, its course was watched through 

 an ordinary balloon theodoUte. When the instrument was detached its fall was followed 

 generally to quite near the ground ; but it was always lost before it actually reached the 

 ground. The direction was then noted in which the instrument was last seen, and the bear- 

 ings taken. With few exceptions this direction was over the sea ice, and then one walked 

 straight ahead in this direction and by the aid of a powerful pair of field glasses the small 

 red rubber balloon was generally seen and the instrument recovered. The method was used 

 on the occasion of the highest ascent about which the following is taken from the note 

 book : — 



' Monday, December 25th, 1911. — ^An attempt was made to reach as high as possible. 

 A fuse for forty minutes was attached. Besides the small red balloon a frame 

 carrying foiu- sheets of silver paper was attached to the instrument. The instrument 

 was detached 43 minutes after the ascent ; but for some time before this the 

 silver paper, and at times the instrument, was all that could be seen below the 

 large balloon. When the instrument fell its course was followed by the flashing 

 of the silver paper. The latter however slowly broke away, and when about 15° 

 above the horizon became invisible. The theodolite was lowered and a bearing of 

 the direction was taken, but as this cut right across Inaccessible Isknd it was 

 not a good one. On setting out to find the instmment I travelled on ski for 

 2 J hours as near the bearing as possible, then I saw the small red balloon through 

 glasses about a mile and half away. The instrument was safely picked up about 

 ten miles from Cape Evans in a S.S.W. direction.' (The instrument had risen 

 6,743 metres.) 



Obviously this method could not be used to reach even moderate heights if there was 

 an appreciable wind. Hence it was only tried on days when Erebus smoke showed that the 

 air was nearly calm throughout the lower atmosphere. 



This method was quite impossible in the winter because in the great cold the moisture 

 from the eye crystallised on to the eyepiece of the theodohte making it unusable. To meet 

 this difficulty a number of cones of fine silk thread were taken south. On each cone was 

 five miles of silk thread the total weight of which was only 4 oz. It was intended to attach 

 one end of this silk thread to the instrument before liberating the balloon. The instrument 

 was then to be detached from the balloon after a given time by means of a fuse, and 

 the thread followed until the instrument was picked up. Naturally difficulties were encoun- 

 tered, but they were of an en.tirely unexpected nature. A simple method for liberating the 

 balloon and paying out the silk without breakuig it was developed, and on most occasions 

 there was little doubt that all went well and the silk thread was intact until the instrument 

 was detached. But occasions occurred with annoying frequency when on following along the 



