METEOROLOGY. 393 



mometer, and this will also be the temperature of the resistance coil. 

 To avoid the error, which would be otherwise introduced by the leads 

 of the resistance coil, the cable was constructed of a double core of in- 

 sulated copper wire, protected by twisted galvanized steel wire. One 

 of the copper cores was connected to each arm of the bridge, and the 

 steel wire served as the return earth connection for both. Sir W. Thom- 

 son's marine galvanometer with a mirror and scale was employed to 

 determine the balance bridge. (Nature, xxvi, p. 190.) 



Commander Magraphi has invented an improved form of iN'egretti 

 and Zambra's deep-sea thermometer, the so-called upset thermometer. 

 In this form the straight thermometer tube is retained in jjlace of the 

 original siphon shape. The revolving fan of the depth meter ceases its 

 action the moment the thermometer is upset. The act of hauling up 

 is that which stops the fan and upsets the thermometer. Any number 

 of thermometers may be fastened upon the sounding line, and all make 

 their record at the same moment when their line is hauled up. {Nature, 

 XXVI, p. 15.) 



Professor Tait has made an exhaustive examination of the corrections 

 to the deep-sea protected thermometers furnished by Casella to the 

 Challenger exploring expedition. These were not wholly protected 

 according to the plan proposed by Sir William Thomson, and were 

 subject to larger errors than was necessary. Tait says corrections of 

 a half degree Fahrenheit per mile may seriously aliect our theories of 

 oceanic water circulation, and demands minute investigation. The 

 Challenger thermometers were all of the Six pattern, and had both 

 maximum and minimum indices, set by means of magnets. Tait also in- 

 vestigated two wholly protected thermometers belonging to Sir C. Wy- 

 ville Thomson, but whose indices are not easily set for recording. His 

 testing apparatus consisted principally of a Frazer cannon, whose avail- 

 able interior was 4^ inches diameter and 4 feet long, affording abundant 

 room to experiment on several thermometers at once. 



The effect of pressure upon ordinary thermometer tubes is very appre- 

 ciable, notwithstanding the relative thickness of the glass sides. A very 

 interesting study of the distortion due to pressure is given by him. The 

 compression is the same for every portion of the glass tube, but it is 

 accompanied by a shear that increases towards the inner surface of the 

 tube until the resulting extension disrupts that surface. When a tube 

 is exposed to dilation by force from within, the dilation of the walls 

 aids the shear. A thin tube is much better able to resist external than 

 internal pressure. {Nature, xxv, p. 90.) 



All the thermometers which have large aneurisms have had special 

 calculations made for them, but in no case does the correction to be 

 applied to the minimum index exceed 0.14° F., or about one-seventh of 

 a degree per mile of depth. The appendices to Professor Tait's report 

 give every detail of his formulae and observations, and afford data for 



