504 REPORTS ON THE STATE OF SCIENCE. — 19] fi. 



at sea sufficiently constant for most experimental purposes. If tliesc 

 experiments possessed no other value they would be still useful from the 

 demonstration of this result. One may congratulate Messrs. Alfred 

 Holt also upon it. 



The outbreak of the war occasioned the transfer of the apparatus to the 

 refrigerator of the P. & 0. R.M.S. Morea (see Interim Report). It was 

 difficult to instal the apparatus, as the refrigerator was only 14 feet above 

 the keel and 12 feet below sea-level, and was approached by three narrow 

 ladders,, but, thanks to the care of Mr. Charlewood and his mate (butchers' 

 department), this was safely accomplished. The writer obtained per- 

 mission to partition a good space of the handling-room, but as it was 

 done with matchboarding it did no more tlian isolate it from the hang- 

 ing joints of meat and other articles which are more appreciated on the 

 upper decks than in the bowels of the ship. At midnight, alone in these 

 depths, on a rough night, with carcases waving to and fro in the light 

 of a rubv lamp (some of the other apparatus was photographic), a whizz- 

 ing fan blowing a blast of snow and air, and the floor frozen and slippery, 

 the conditions were not those to be deliberately sought by a scientific 

 investigator 



The chambers on the Morea were cooled by air wliich was blown into 

 and extracted from them. The same facilities for maintaining a constant 

 temperature were not available, and so the two systems cannot be pro- 

 perly compared. 1 think, though, that the brine system is more satis- 

 factory and more rapid in compensating for the introduction of the 

 observer. It. was found preferable on s?. Asmiiius to introduce the brine 

 at a temperature as little below that required by tlie room as possible. 



As the engineers on R.M.S. Morea did not find it practicable to run 

 the engines more than twice or three times a day, t arranged my fan in 

 such a way that it sucked from a neighbouring cold-chamber a quantity 

 of cold air which it was hoped would compensate for my entry, but it 

 did not make much difference. 



It will be seen from fig. 16 that the conditions in the experimental 

 chamber on the homeward voyage were much less favourable as regards 

 temperature. 



12. Influence of Grainty Deviations iifon Meteorological Phenomom. 



Though the effect is necessarily very small, it is just possible that 

 under some conditions the influence of variations in gravity upon 

 meteorological conditions may prove appreciable. For example, the 

 change in the value of g, which a body experiences when it moves E. 

 or W., will apply to the motion of a mass of air. A current going east 

 {i.e. a westerly wind), being attiacted less, tends to rise, whereas an 

 easterly wmd tends to d-^scend. A velocity of 13'7 knots per hour in an E. 

 or W. direction at the Equator is equivalent to a change of barometric 

 pressure equal to O'l mb. 



There is a considerable difierence between the gravitational attrac- 

 tion upon a mass of air moving N. or S. according to whether it assumes 

 the velocity of the earth below it or not ; for example, the decrease in 

 gravity for motion from Lat. 50° to Lat. 45° is equivalent to a change 

 of pressure of nearly 0'5 mb. 



Then, again, any wide departure from isostatic eqtuUbrium, such as 

 is suggested by this research over the Indian Ocean, may show itself over 



