August 30, 1S83] 



NA TURE 



4i9 



into the air, allowed to drain for some time in a tempera- 

 ture of 32 or upwards, and then be washed in fresh 

 water, it will be found to be nearly quite free from salt, 

 and the water produced from it may be drunk." 



During the Antarctic cruise of the Challenger the writer 

 of this notice made some experiments to decide the 

 question whether or not sea-water ice is a mixture of pure 

 ice and sea water or brine. The melting point of salt- 

 water ice of various sources was carefully observed, with 

 the following results. Ice formed in a bucket of sea water 

 over night melted at - 1 '3° C. The bulk of ice formed was 

 insignificant compared to the volume of water in which it 

 was formed, so that this was a specimen of bona fide sea- 

 water ice, without admixture of snow or spray. In the 

 same way the melting point of pack-ice was determined. 

 The freshly collected ice began to melt at - i° C. ; after 

 twenty minutes the thermometer had risen to -o'0°, and 

 two hours and a half afterwards it stjod at -o'3°, having 

 remained constant for about an hour at -0"4°. -Another 

 portion of the ice rose more rapidly in temperature, and 

 when three-fourths of the ice was melteJ, the thermo- 

 meter stood at o° C. In the case of the ice frozen in the 

 bucket, the melting point remained constant for twenty 

 minutes at - I - 3°, after which no observations were made, 

 so that we do not know if this ice, formed under the most 

 favourable circumstances, showed the same irregularities 

 as the pack-ice, picked up out of the sea ; but as the bulk 

 of ice experimented on did not much exceed 10 cubic 

 centimetres, the greater part of it must have melted in 

 the twenty minutes. Indeed as the amount of ice formed 

 in the bucket did not sensibly alter the composition of 

 the water left liquid, there seems to be no reason why 

 the ice should not be a homogeneous substance. 



Adhering brine can have no influence on the melting 

 point of ice, consequently, if sea-water ice consists of 

 pure ice with entangled brine, it must melt at o' C. If 

 its melting point is different from o° C. then the solid 

 matter of the ice is not pure ice. We have seen that 

 frozen sea water has a melting point of - i°-3, which 

 .s fairly constant, and that pack-ice, which must neces- 

 'sarily be formed by the freezing of salt water, the con- 

 gealing of spray, and the accumulation of snow, begins to 

 melt about - i°, the temperature gradually rising as the 

 constituents of lower melting point are liquefied. It is 

 thus readily apparent how it is that Scoresby found that 

 such ice " allowed to drain for some time in a temperature 

 of 32 or upwards," produced in the end potable water. 

 The salt-water ice of low melting point effectually pre- 

 vents the intermingled snow from melting, which finally 

 remains practically intact, and of course can be drunk on 

 being melted. 



Dr. Petterson on purely chemical grounds comes to the 

 same conclusion. He says (p. 303) : " Those who support 

 the common theory that sea ice is in itself wholly desti- 

 tute of salt, and only mechanically incloses a certain 

 quantity of unfrozen and concentrated sea water, must 

 confess that we in this case ought to find by chemical 

 analysis exactly the same proportion between CI, MgO, 

 Cao, S0 3 , &c , in the ice and in the brine as in the sea 

 water itself." That this is not the case is shown by a 

 number of analyses of sea-water ices in which the propor- 

 tion of CI: S( .' 3 varied from 100 : 12S to 100: 766, the 

 average proportion in sea water being 100: 1188. The 

 results of his investigations may be summarised as 

 follows : — 



C'cean water is divided by freezing into two saliniferous 

 parts, one liquid and one solid, which are of different 

 chemical compositions. Taking the relation CI : St> 3 as 

 standard of comparison, the most striking feature of the 

 freezing process is that the ice is richer in sulphates, and 

 the brine in chlorides. The extraordinary variation, both 

 in saltness and in chemical composition of every indi- 

 vidual specimen of sea ice and sea brine, shown by the 

 tables, depends on a secondary process, by which the ice 



seems to give up its chlorides more and more, but to 

 retain its sulphates. Hence the percentage of chlorine is 

 no indication of the saltness of the ice, though it may to 

 a certain extent be taken as an index of its age. 



In connection with this part of the subject, the author 

 cites Prof. Guthrie's work on Cryohydrates, and gives 

 the following table : — 



Cryohydrate of Contains Solidities at 



NaCl ... 76 "39 per cent, water ... -22" C. 



KC1 ... 8000 ,, ,, ... - u° - 4 C. 



CaClj ... 7200 „ „ ... -37 C. 



MgS0 4 ... 78-14 ,, „ ... - s°oC. 



NajS0 4 ... 95-45 „ „ ... - 0=7 C. 



Supposing that these cryohydrates are formed in the 

 freezing of sea water, it is easy to see how, as the tem- 

 perature rises, the chlorides melt out first and leave the 

 ice richer and richer in sulphates. 



Before concluding this notice, attention must be called 

 to a statement in a note at the foot of p. 318 : "As a 

 thermometer immersed in a mixture of snow and sea 

 water, which is constantly stirred, indicates — i 3, 8 C, we 

 may regard," ecc. This can be true only if the tempera- 

 ture of the atmosphere is — i°'8 C. ; if it is o° C. or 

 higher, the temperature of the sea water will assuredly 

 rise to the melting temperature of snow, or o" C. 



Even though it should turn out that chemically pure 

 ice does, as the author suspects, melt suddenly without 

 previous contraction as ice, the discovery of the existence 

 of a minimum density point of ice, not chemically pure, 

 which includes all the ice on the globe, is one of the very 

 highest importance. 



It is to be hoped that we shall soon have a further 

 instalment of work on a subject so large and so important, 

 and with which the author has shown himself so well 

 qualified to deal. J. Y. BUCHANAN 



THE STABILITY OF MERCHANT 

 STEAMSHIPS 



T PROPOSE to state, and in part to restate, the more 

 •*■ important scientific considerations concerning the 

 stability of merchant steamships which the investigation 

 of the Daphne disaster has brought to light, following the 

 main lines of the second part of my Repot t, which has 

 been published in extenso in several newspapers. In this 

 case, as in all cases touching the complicated question of 

 ship stability, it is very necessary to be careful not to 

 draw hasty inferences or any inferences at all which are 

 not strictly deducible from the facts or principles esta- 

 blished. 



It is desirable to guard the reader in the first place 

 against considering the cases of the ships Daphne and 

 //ammonia — which I have had occasion to associate 

 somewhat closely in my Report — as identical in more 

 than a certain number of features, there being other fea- 

 tures in respect of which there is little or no resemblance. 

 I will presently point out both the resemblances and the 

 differences, but first let me remind the reader unfamiliar 

 with naval science what is meant by a curve of stability, 

 quoting the Report as far as may be necessary for the 

 purpose. Fig. 1 may be taken as the transverse section 

 of a vessel inclined at an angle of 15 degrees to the up- 

 right. The total weight or gravity of the vessel will act 

 downwards through the centre of gravity G, and the total 

 buoyancy will act upwards through the centre of buoyancy 

 B, as the arrows indicate. It will be obvious that the 

 vessel cannot rest in the inclined position with these forces 

 and no other operating upon her ; she must revolve until 

 gravity and buoyancy act in the same vertical line, but in 

 opposite directions. The further she is inclined the more 

 will the ship be immersed on one side and emersed on 

 the other, and therefore the further out will the centre of 

 buoyancy move. Now as neither the gravity nor the 

 buoyancy need be altered in amount by mere inclination, 



