In a recent paper we drew attention to the variations of cosmic ray in
tensity from place to place within 1 kpc of the position of the sun. A
consequence is that the use of the 'gamma ray technique' to determine
the masses of molecular clouds ceases to be as straightforward as pre
viously thought, Here we examine the problem in some detail by making
various assumptions about the manner of cosmic ray variation and concl
ude with a 'best-estimate' of the conversion factor X (= 2N(H-2)/W(CO)
) between the column density of molecular hydrogen, N(H-2) molecules c
m-2 and the line integral of the (CO)-C-12 J = 1 --> J = 0 transition
over velocity. The average value, for local clouds within 1 kpc, is [X
] = (1.7 +/- 0.4)10(20) mol cm-2 K-1 km-1 s.