284 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
who work in London and what they have done, we must recognize 
the fact that even here the conditions are not so bad as we sometimes 
imagine. I have often been strongly impressed (since my work in 
Chicago) with the belief that a smoky atmosphere has some advan- 
tages in astronomical work, for it seems that the seeing is frequently 
improved in solar observations when the sky is smoky. Here is a 
fine chance to test that question, and I think it has been tested at 
Greenwich, and that some of the photographs taken there (both 
solar and stellar) prove that London smoke does not prevent excel- 
lent definition. I examined rather carefully some plates there yes- 
terday, and the star images are surprisingly good in many instances. 
It seems to me that definition by night as well as by day at Green- 
wich must be of an order much higher than one might suppose when 
one thinks of Greenwich as being within the boundaries of London. 
But it is perfectly possible to get good results anywhere, provided 
sufficient care is taken. One must consider, for example, the best 
time of day for solar work. It usually happens that the best defin1- 
tion of the sun occurs in the early morning and the late afternoon. 
Mr. Newall tells me that this is as true at Cambridge as it is at Mount 
Wilson. This is worth looking into if one takes up work on the sun. 
Further, one must have a definite plan of work. This is of prime 
importance. Devote your entire attention to a single investigation, 
involving, if possible, two or three parallel series of observations, so 
devised as to throw light on one another. Frequently the value of a 
given series of observations may be enormously enhanced if other 
observations are available to aid in their interpretation. For exam- 
ple, in studying the spectra of sun spots, the character of the spots, 
their motions, and changes of form, and the distribution of the 
flocculi in their neighborhood, may be vital factors in interpreting 
the spectroscopic phenomena. Then, again, there is the great possi- 
bility that new methods and new instruments can be applied. Up 
to the present time I think the interferometers of Michelson or of 
Pérot and Fabry have never been systematically employed for work 
on the sun. That admirable method which Fabry is using in the de- 
termination of absolute wave lengths would perhaps be very useful 
indeed if applied to the measurement of the displacement of solar 
lines at the center and at the limb. I also believe that the echelon 
spectroscope has never been used for the observation of the narrow 
bright lines in the chromosphere. Furthermore, we are always con- 
fronted by the possibility of perfecting our optical apparatus. I have 
been trying for years te get good prisms of large size, but can not get 
homogenous glass, and therefore it now seems necessary to attack the 
@While revising this for republication I learn with great pleasure of the dis- 
covery at Greenwich of an eighth satellite of Jupiter.—G. HE. H. 
