( 644 ) 



used tubes with two openings (see fig. 6 PI. I) through wliich the 

 two wires of tlie element are drawn, yet in many of these tubes 

 tensions apjieared, wliieh jiroved an impediment to their being- 

 operated upon. 



Formerly it was very diftieult to make a eonnection between the 

 C'opperblock (for the meaning of this ef. Comm. N".27) and the protecting- 

 tube, which connection should not only be airtight, but which also 

 must allow of l>eing placed into steam and at low temperatures into 

 different kinds of liquids. This difficulty has been entirely overcome. 

 In N". 27 we have spoken of our intention to try, following the method 

 of Cailletet, and solder the copper block on to glass which to this 

 end had been platinized. In this we succeeded to perfection. The 

 glass tube is platinized at the end with platinum-chloride in the blow- 

 pipe while care must be taken that it remains perfectly cylindrical, 

 then it is coppered galvanically and tiniied over an alcohol flame. 

 Then the thin upright rim turned on to the block, which is also 

 tinned and carefully cleaned, is pushed round the end after wdiich 

 they are soldered together Iw means of resin as a tlux. Then the 

 cap, the juncture seam and the timied glass together are galva- 

 nically platijiized and gilt. In this manner an important improvement 

 has been obtained. The indiarubber protecting-ring which made the 

 limb much thicker is removed, the fit is perfectly tight and permanent, 

 the limb may be fastened into almost any api)aratus and be inmiersed 

 into any liquid Avithout the least fear of action on the wires or of 

 shunt-circuits anywhere between the two wires. 



Platinizing and gilding are not ah\ays necessary. The thermo- 

 element used in the experiments of Hollmann 1. c. was only tinned, 

 as acetaldehyd docs not attack tin. If for one reason or other we 

 do not wish to bring the protected thermoelements into direct contact 

 with a liquid with strong chemical action in which it is immersed, 

 they are enclosed in a separate protecting-tube (see figs. 3 and 4, 

 PI. I) terminating in a thinwalled copper cap soldered on to it in 

 the way as described above, into Avhich the block of the thermo- 

 element fits exactly, and which cap is covered galvanically with a 

 suitable metal. This auxiliary means was for instance used here 

 in determinations of melting-points of mixtures of chloride and 

 sulphur by Mr. Aten. Between the limb of the thermoelement and 

 the protecting-tube provided with a platinized and gilt cap, a little 

 pentane was poured to till uj) the space. 



In addition to the description of Comm. N". 27 we remark that 

 the airtight comiection of the outer glass protecting-tubes with the 

 indiarubber protecting-tube has been made by means of indiarubber 



