TRIBE ON FREEZING OF WATER AND BISXUTH. IX-Freezing of Water and Bismuth. By ALFRED TRIBE,F.C.S. PROFESSOR in his work entitled ‘‘ Heat considered TPNDALL as a mode of Xlotion,” after speaking of the anomalous expan- sion of water prior to solidification and the importance of this property in the economy of nature controverts the generally received opinion that no other bodies possess the same pro- perties in this respect and cites bismuth as a case in point. He says at page 84 ‘‘ The fact however is that the case is not an isolated one. You see this iron bottle rent from neck to bottom; I break it with this hammer and you see a core of metal within. This is the metal bismuth which when it was in a molten condition was poured into this bottle the bottle being closed by a screw exactly as ill the case of the water.The metal cooled solidified expwi(Ie,l :tiid the force of its expailsion was sufficient to burst tlic bottle. There are no fish here to be saved still the molten bismuth acts exactly as tlic water acts .7 The experiment as above detailed shows conclusively that molten bismuth expands either on cooling or freezing; but it affords very little or no proof of that ~vliich has been inferred from it. The real question obviously is not whetlier the fluid metal in returning to the solid condition rends an “iron bottle from neck to bottom,” but whether the freezing of water arid bismuth are analogous. The results and details of the experiments which have been made with a view to answer this qnestion form tlie subject of the present communication.To ascertain the general mode of solidification paraffin nitrate chlorate and hydrate of potassium and nitrate of sodium were experimented with. The first experiments were made by simply melting bismuth paraffin &c. in good sized test-tubes. The solidification of the transparent molten substauces on cooling could easily be observed to start from the bottom. The solidification of bis-muth likewise was found to commence at the bottom which TRIBE ON FREEZING OF WATER AND BISMUTH. was ascertained by pouring off the fluid metal after it had somewhat cooled. Water was observed by placing tubes about three-quarters full of the liquid in a freezing mixture-taking care to have the mixture some little distance above the surface of the wat,er.In every experiment solidification began at the surface; but in one especially ice was seen gradually to descend while in the freezing mixture to the bottom. Two or three times during the descent the tube was removed from the freezing mixture which instantly stopped the formation of ice. Thinking the tubes employed favoured the solidification of bismuth &c. at the bottom (there being comparatively a larger surface exposed to the cooling action of the air) the experi- ments were repeated in U-shaped tubes but with the same results. The experiments were likewise made in crucibles again with the same practical results. A slight modification of this form of experiment submitted the metal to it severe and what was considered conclusive test.A crucible three-quarters filled with molten bismuth was placed over a Emall Bunsen’s gas jet and in this way allowed to cool extremely slowly-at the same time cold air was blown upon its surface. It will be seen that this experiment decidedly fiavoured solidification on the surhce ; but the fkeezing of the metal commenced in every caRe at the bottom-the process being repeated several times. At this stage of the inquiry the question presented itself Have the air-currents produced by the heated tubes &c. any-thing practically t,o do with solidification at the bottom. To ascertaiii this tubes filled with melted bismuth and paraffin were placed horizontally in the centre of a cylinder which was then immediately closed.The solidification of the paraffin-it being transparent-could be distinctly seen to start from the bottom. The bismuth was ascertained AS before by pouring off the surface metal. From the foregoing experiments it appears that as a rule substances solidify upwards-water being the exception. Whe-ther water be the only exception it is certain that bismuth follows the rule. With regard to those bodics which commence solidifying at the bottom we are entitled to re;tsoii a8 follows :-As soon as a liquid begins to cool its temperature ceases TRIBE ON FREEZXNG OF WATER AND BISMUTH. to be uniform some particles being more exposed to cooling influences than others. The necessary consequence of this must be a difference in the specific gravity of the different particles which will arrange themselves accordingly.It may be assumed therefore that after a short time the denser particles will be found at the bottom and as solidification begins there the conclusion may be drawn that the point of greatest density coincides with that of solidification and it stands to reason that the particles which solidify first are the coldest. It may be concluded from this that molten bismuth on cooling does not expand before freezing as is the case with water but that it continues to contract down to the solidifying point. The rapid expansion which is well known to all chemists is due to the act of crystallisation. The reason why the metal does not rise to the surface as soon as it becomes solid although specifically lighter is siniply because it adheres to the bottom and sides of the vessel.The two following experiments confirm in a great measure the above deductions Melted paraffin and water at 100" C. were placed in tubes about six inches long closed at both ends with corks and then completely imniersed vertically in cold water for a few minutes. It was found by a thermometer that the top of each liquid was about 3' C. higher than the bottom. Hence the analogy of bismuth with water is not perfect but fails just in that point which causes our rivers and lakes to freeze from the surface and thus protect animal life. The above experiments were principally through the kind- ness of Dr. Gladstone made in his laboratory and I am much indebted to him for many valuable suggestions.