1 62 CARNEGIE INSTITUTION 



certain spectrum given us by a spot of area 200 at minimum and 

 another spectrum given us by one of 2,000 at maximum, to ascribe 

 the change to a quality in the period (if I may so express it) when 

 it may be a function of the size of the spot itself. Further, the 

 average spot group goes through a certain pretty well defined routine 

 in the course of its growth and decay. Now, if we are at the mini- 

 mum of the cycle, our groups run small, and it is only (in most 

 instances) during one particular phase of its development that a 

 group is likely to be a tempting object for spectroscopic examination. 

 At maximum we may have plenty of giant groups, which can easily 

 be followed spectroscopically during their whole career. Here, again, 

 is a point which wants to be followed out. If we record a given 

 spectrum for a certain spot, we have not learnt all that we can unless 

 we trace the history of that spot back to its rise and onward to its 

 disappearance, and determine at what particular stage of its develop- 

 ment the observation was made. 



We want to know whether we can associate different spectroscopic 

 appearances with — 



(i) The size of the spot ; since the larger spots may be assumed 

 on the average to be the deeper. 



(2) The stage of its development. The depth may alter with the 

 age. 



(3) The changes that are going on in the group. 



(4) The progress of the general solar cycle. 



(5) The type of the spot group. 



I have put the type last, out of its logical order, because it has 

 seemed to me that, though the great majority of spot groups con- 

 form to one general type of evolution, yet occasionally we get spots 

 of a very distinguishable form, and it is these spots, when of im- 

 mense size, and not spots of the normal type, that are clearly and 

 unmistakably associated with magnetic storms. It would be no 

 small matter if we found that such spots exhibited some distinct 

 spectroscopic peculiarity. 



As to the method of observation, clearly the photographic regis- 

 tration of spot spectra should be the routine one ; but it certainly 

 should not exclude the direct visual work. Just precisely as our 

 daily photographs of the sun's surface at Greenwich, however ad- 

 mirable for their purpose, leave us without any record of the pro- 

 cesses of rapid change, so it would be with the photographic regis- 

 tration of sun spot spectra. They cannot possibly render direct 

 work unnecessary. 



