124 Prof. E. Wiedemann on a Means to determine the 



which interference is possible. According to the last-men- 

 tioned observer, a difference in path of 50,000, and even 100,000 

 wave-lengths still produces appreciable interference-bands in 

 sodium light. The time corresponding to this difference of 

 phase is about 0'5 x 10 -10 second. Calculating approximately 

 the mean free path (as we only want to compare the orders of 

 magnitude) for hydrogen at 0° and atmospheric pressure, we 

 find for the time necessary to traverse this free path 1*14 x 10~ 10 

 second. The two numbers are sufficiently close to justify the 

 assumption that the sodium atoms may vibrate during 50,000 

 oscillations without sudden change of phase. We see at the 

 same time that the disappearance of Newton's bands need not 

 be due to the widening of the lines producing them. 



As the mean time elapsing between two encounters depends 

 chiefly on the pressure, and far less on the temperature, we 

 have a means of determining approximately the pressure of a 

 gas by an examination of the light which it sends out. 



The determination of the quantities relating to large differ- 

 ence of paths in the light sent out by the sun, its protuberances, 

 and stars may give us important information on their physical 

 constitution ; and I should like to draw the attention of spec- 

 troscopists to this point, now that we may soon expect a re- 

 newal of the sunspot maximum. 



In order to make the measurements, we need only decom- 

 pose the light we want to examine by means of a spectroscope, 

 separate a ray, which must be as homogeneous as possible, and 

 count the number of Newton's rings visible between two ad- 

 justable pieces of glass. We might also determine the thick- 

 ness of a plate of Iceland spar which still shows interference- 

 bands if it is placed between two Nicol's prisms. The plate 

 must be cut parallel to the axis; and particular attention must 

 be paid to its homogeneousness. 



The spectroscopic evidence hitherto has only related to the 

 presence of a substance in the sun; and we only derive from it 

 very general notions as to the physical state. The radiation 

 of heat is, as Janssen has recently again had occasion to 

 observe, a very complicated phenomenon. We must take 

 account of all the different layers on the sum's surface ; and 

 the same difficulty besets the interpretation of the Fraunhofer 

 lines. Their thickness and darkness is a function of at least 

 three variables — temperature, pressure, and thickness of ab- 

 sorbing layer ; only after having successfully investigated the 

 separate effect of these three variables, will a more perfect in- 

 terpretation of the phenomena be possible. 



If electric phenomena are going on at the solar surface, 

 the difficulty of the ordinary methods of investigation will be 

 still further increased ; for, as I have shown, 



