The composite galaxy luminosity function ( hereafter LF ) of 39 Abell clusters of galaxies is derived by computing the statistical excess of galaxy counts in the cluster direction with respect to control fields .
Due to the wide field coverage of the digitised POSS-II plates , we can measure field counts around each cluster in a fully homogeneous way .
Furthermore , the availability of virtually unlimited sky coverage allows us to directly compute the LF errors without having to rely on the estimated variance of the background .
The wide field coverage also allows us to derive the LF of the whole cluster , including galaxies located in the cluster outskirts .
The global composite LF has a slope \alpha \sim - 1.1 \pm 0.2 with minor variations from blue to red filters , and M ^ { * } \sim - 21.7 , -22.2 , -22.4 mag ( H _ { 0 } = 50 km s ^ { -1 } Mpc ^ { -1 } ) in g,r and i filters , respectively ( errors are detailed in the text ) .
These results are in quite good agreement with several previous determinations and in particular with the LF determined for the inner region of a largely overlapping set of clusters , but derived making use of a completely different method for background subtraction .
The similarity of the two LFs suggests the existence of minor differences between the LF in the cluster outskirts and in the central region , or a negligible contribution of galaxies in the cluster outskirts to the global LF .