NO. 8.] PENDULUM OBSERVATIONS. 7 



The above observations thus give the following values for the tempera- 

 ture, (the reduction for Tonnelot to the hydrogen-thermometer being 0.095 

 at this temperature): 



according to Tonnelot 4519 Pendulum therm. 20 Pendulum therm. 23 

 15.852 C. 15.916 C. 16M13 G. 



Thus, with Von Sterneck's formulae, both the pendulum thermometers give 

 somewhat too high a temperature, indicating that the zero has risen. 



I have examined thermometer 20 repeatedly. The year after 1 received 

 the pendulum apparatus, in the summer and winter of 1893, some compari- 

 sons were made in the air with a thermometer Baudin 8967, of which the 

 corrections to the hydrogen-thermometer were determined by the aid of a 

 thermometer Tonnelot 4506, which had been examined at the "Bureau Inter- 

 national des Poids et Mesures". I have since made several series of compa- 

 risons, some with Baudin 8967, some with Tonnelot 4519, with the two ther- 

 mometers hung up side by side in water. It appears that thermometer 20 

 has not changed more than a couple of hundredths of a degree since the 

 summer of 1893, and that the zero has risen about 0.09. My temperature 

 comparisons give the following formula: 



* C. = 1-863 (reading 3-46). 



It is the same with the other thermometer, 17, belonging to the pendulum 

 apparatus for the "Gradmaalings Kommission". Its zero has risen 0.12 G. 

 I assume therefore, that it may be taken for granted that the rise of the 

 zero in the case of thermometer 23 had already taken place before the ob- 

 servations made during the expedition were begun on July 30th, 1893. 

 According to the above observation, I have assumed a rise of 0.25 in the 

 following calculations, and in so doing will remark that a more exact agree- 

 ment between the indications of the pendulum thermometers than within a 

 few hundredths of a degree cannot be expected, as one division on them 

 answers to 0.2, and the thermometers are not calibrated. The second ther- 

 mometer of the expedition, 24, shows a zero-rise of 0.375. 



The following are the observations made, in Vienna, in Ghristiania, and 

 during the expedition. For the calculation of the time of coincidence, c, we 

 have, by the method of least squares, 



