236 Intelligence and Miscellaneous Articles. 



still more surprising and remarkable. If such a one by careful treat- 

 ment is cooled down to 0°, and then made to crystallize suddenly, 

 the crystal cake formed, which constitutes a compact solid mass, 

 exhibits an extraordinary increase in volume, and on further cooling-, 

 to about 10° C. below zero, contracts more and more. As in this 

 condition of the original solution there can be no question of a sepa- 

 ration of crystals as in the former case, it seems (like water below 

 4°) not to follow the law according to which bodies contract by di- 

 minution of temperature. 



The experiments in question, which any one can easily execute, I 

 made with a glass flask of about 60 cubic centims. capacity, into 

 which in different experiments I poured solutions of Glauber's salt of 

 various strengths, covered them with a layer of petroleum, and closed 

 the flask with a perforated caoutchouc stopper. Through this stopper 

 passed a glass tube of 30 centims. in length and 2*09 cubic centims. 

 in capacity, provided with a paper scale and reaching down to the 

 layer of oil. To vary the temperature, the flask could be placed in 

 a beaker containing either a freezing-mixture or warm water. If 

 the crystallization of the enclosed solution did not take place spon- 

 taneously at the proper moment, it was immediately produced by a 

 small particle of crystal projected through the open tube. The 

 sudden change of temperature was frequently so energetic that 

 the flask was cracked if the glass tube was accidentally stopped 

 and presented no outlet for the displaced covering layer. The 

 whole apparatus could be used as a thermometer (only in an inverse 

 manner), and indicated a change of temperature of not too brief 

 duration in a tolerably delicate manner. If, for instance, the strongly 

 cooled apparatus was warmed with the hand for a few moments, a 

 depression of the covering layer was distinctly perceived. — Poggen- 

 dbrff's Annalen, May 1866. 



ON THE FIGURE OF THE EARTH. BY CAPTAIN. A. R. CLARKE. 



To the Editors of the Philosophical Magazine and Journal. 



Ordnance Survey Office, Southampton, 

 Gentlemen, August 17, 1866. 



The question of the correctness of the simple and direct applica- 

 tion of the method of least squares to the determination of the figure 

 of the earth, which is controverted by Archdeacon Pratt, is an im- 

 portant one, but, being purely mathematical, should not be a matter 

 of opinion. As, however, it cannot be of great interest to the ma- 

 jority of your readers, I shall only ask leave to remark here that, so 

 far from having vindicated the legitimacy of his " correction " of the 

 method of least squares, his "improvement upon his correction" is 

 not only expressly hypothetical and arbitrary, but is a still further 

 departure from simplicity and truth. The values which Archdeacon 

 Pratt, in his work ' On the Figure of the Earth,' had obtained for 

 the local attraction at the " reference-station " of the three great arcs 



