280 REPORT— 1880. 



lines was taken by Lockyer for a reversal, and that consequently the greater 

 intensity and width of the more refrangible line appeared as a one-sided 

 development of the wings. The second case is not very clearly described 

 by Ciamician, In the text of the paper he only mentions that when the 

 double yellow mercury line becomes fuzzy, continuous light is seen to the 

 right and left of it. In the drawing this is figured as a widening of 

 the more refrangible line towards the violet, and of the less refi-angible 

 line towards the red. The description is too vague to allow any certain 

 inference to be drawn, but it seems possible, as a simple optical fact, that, 

 when a double line widens, the wings can be traced to a greater distance 

 on that side of each component which is removed from the other. 



We have now to discuss the causes which may produce the widening 

 of lines. In the first place it might be suggested that, in accordance 

 with a formula which we have ah'eady given, an increase of the quantity 

 of luminous matter would produce an apparent widening of the lines ; 

 for it follows from the formula that, unless the coefiicient of absorption 

 is absolutely zero for any given wave-length, the spectrum sent out by an 

 infinite number of molecules in the line of sight is always continuous. A 

 greater number of molecules will, therefore, cause the spectrum to ap- 

 proach the continuous state, and the widening of the lines may be due to 

 the first stage towards this approach. To this wo shall reply, that an 

 increase in the number of molecules cannot be the primary cause of the 

 widening of lines ; for the lines of sodium, for instance, in the sun are 

 compai'atively sharp, though the thickness of the absorbing layer is 

 greater than anything we can produce in our experiments. We can 

 prove the same point more clearly in the case of hydrogen. If we enclose 

 the gas in a tube of the form adopted by Mr. Monkhoven and Prof. 

 Piazzi Smyth, so that we may look longitudinally through the capillary 

 bore, we increase the thickness of the radiating layer to a very great 

 extent ; yet lines which are sharp when the tube is looked at transversely, 

 will remain sharp when it is looked through longitudinally, although an 

 increase in the pressure or in the intensity of the discharge will at once 

 produce a widening. It must be remembered that the eS'ect of an in- 

 creased number of radiating molecules will only depend on the curve of 

 intensity near the line. If a line is absolutely sharp, no increase in the 

 number of molecules will ever increase its width, and two lines of the 

 same brightness, which present the same appearance at their edges, must 

 behave exactly in the same way, when the thickness of the radiating layer 

 is increased. Yet, while we have some lines which widen easily and 

 enormously, others, which present the same appearance, do not show any 

 widening. That, if a line is once widened, an increase in the number of 

 radiating or absorbing molecules will increase the apparent extent of the 

 widening is possible, but we must distinguish this effect from the original 

 cause which bas produced the widening. 



It is, I believe, the almost unanimous opinion of spectroscopists that 

 the widening is, in most cases, produced by an increase of pressure. 

 This opinion was first put forward by Frankland and Lockyer.' In the 

 case of gases, the easiest way to produce the widening is by an increase 

 in the pressure of the gas, and the metallic lines are also generally seen 

 to widen when the density of the gas through which the spark is taken 

 is increased. But we may also, in the case of hydrogen, for instance, 



» Proc. Boy. Soc. xvii. p. 288 (1869). 



