60 
the resulting values of the magnitude of R for the night 
being therefore 9*7, 9*7, 9*7. It is difficult to imagine I 
could have made an error of half a magnitude here ! The 
descending curve is, you will see, very regular. It seems 
to me that this is rather an interesting point in variable star 
observations, and worthy of some study.” 
The projection of the observations of R Vulpeculse 
referred to in this extract was exhibited to the meeting, 
and the general impression of the members present seemed 
to be that the irregularity in the ascending part of the curve 
could not fairly be attributed to errors of observation. 
