( 374 ) 



By subtracting these annual means for eaeli limb as they are 

 given in the second and third columns of' the table above from Ihe 

 means foi- the same limb h)r each daj' of the moon's age I obtained 

 for each jear a set of about 25 residuals and finally I combined 

 the corresponding residnals of the 8 years with their i-espective 

 weights. 



These mean residuals follow here : 



If we assume the adopted values for the inequalities depending 

 on D to be correct, the two preceding tables show us the effect 

 of the systematic errors of the obserx^ations. At a first glance at the 

 second table we perceive that the right ascensions observed at the 

 age of 4 and 26 days show abnormally great discordances, which 

 both agree in sign with those which would result if the observers 

 estimated the moon's diameter to be smaller when observed at 

 daylight. 



If we except these two groups, the observations of the 2"^ limb 

 no longer show any regular increase, whilst the results for the 

 1^^ limb between the ages of 5 and 8 days still seem to vary somewhat 

 regularly. However, after due consideration of the case, I have ulti- 

 mately assumed the personal error to be constant for the first limb 



