or other is not so intense or so extensive in area, it means that there is a breakdown of the north-east monsoons in India, and a breakdown of the north-easterly system in the Barents Sea. Hence, we have this further relationship, that if there is a breakdown of north-east monsoons in India, there is a minimum amount of pack ice in the Barents Sea and on the shores of Spitsbergen, which reminds one of Mossman's dictum, that the failure of the winter rains on the coast of Chile, south of lat. 33°, means that the Weddell Sea is comparatively clear of ice.
It must be emphasised that well-trained meteorologists are essential for conducting thoroughly satisfactory observations, for there are many errors that unguided amateurs are apt to commit, however conscientious they may be in the task set them.
The selection and setting up of instruments, either on board ship or ashore, is important. Before the departure of the Scotia I was aware that temperature observations on board ship were often vitiated by the warmth from the ship itself according to the relative direction of the wind. Yet, in spite of this well-known fact, I have not known any other ship but the Scotia fitted out with a double set of thermometers, one on the port side and the other to starboard. This was the arrangement
