572 Transactions. — Chemistry and Physics. 



from a study of these logs of twenty-three voyages in the 

 Southern Ocean during the last eight years. There are, how- 

 ever, other facts which have considerable value, such as the 

 ordinary rate of eastward progress of the cyclonic storms, 

 which, as I before mentioned, is about ten or twelve knots an 

 hour, although it is occasionally accelerated or retarded, and 

 sometimes is deflected northwards or southwards by the in- 

 fluence, I believe, mainly of anticyclones. 



On only one occasion (in April, 1892) is the fact mentioned 

 that observations had been made on the temperatures of the 

 air and the sea in the vicinity of icebergs. Mr. Barthorp, 

 however, informs me that such observations have been regu- 

 larly made, and with the disappointing result that no reliable 

 information evidencing the proximity of an iceberg is to be 

 obtained by thermometrical observations, but that in the dark, 

 and in foggy and thick weather, the only means available to 

 safeguard a ship against dangerous collisions with icebergs is 

 to keep always a most vigilant look-out. 



Assuming that the main cause of high and low barometer 

 readings at the level of the sea is, as I have given strong 

 reasons for believing, downward or upward motions of the air 

 in its vertical circulations, the ships' logs under consideration 

 seem to indicate a normal downward motion between latitude 

 30° and 40° south, upward about 50° south, and again down- 

 ward about 60° south, and it would appear that westerly 

 counter-trade winds usually felt at about 40° to 50° south are 

 really the northerly parts of a belt of cyclones, and that when 

 a ship pursues a more southerly course than usual, or when 

 the belts of cyclones and anticyclones are in more northerly 

 latitudes than usual, then easterly winds are experienced, 

 either resulting from the circulation of the southern parts of 

 cyclones or the northern parts of far south anticyclones. This 

 very probable interpretation of the facts observed during the 

 numerous voyages under consideration is, of course, more im- 

 portant to sailing-ships than to steamers ; but it may be 

 useful to either, and from a scientific point of view it leads us 

 to conclude that the vertical circulation of the atmosphere 

 between the equator and the poles, although in principle that 

 shown in the diagram attached to my paper on the subject 

 which appeared in the Proceedings of the New Zealand 

 Institute for 1889, is probably somewhat more complicated 

 in southern regions, and of the character indicated in the 

 revised diagram hereto annexed (see Plate LV.). 



It may be, however, apart from cyclonic circulation, that 

 there are other special causes producing very low pressure in 

 antarctic regions. 



Pig. 1 gives a half-section of the Southern Hemisphere, 

 showing on its edge the probable vertical system of circula- 



