The following information is from a letter dated
5 June 1987 from R.C. Lilwall to R.E. Habermann

Contact R.C. Lilwall if you have any questions
concerning these data.
------------------------------------------------

Thank you for your letter of the 19 May concerning
maximum-likelihood mb magnitudes. Enclosed is our
current earthquake data tape which contains my
most recent mb determinations covering the period
1964-83 inclusive. Attached are details of the
data format for each file and a suggested FORTRAN
format to be used to extract the data. The
following notes may be helpful.

1) The tape covers the period 1964-87 but more
   recent (Post 83) data have no max-like values.
   Apart from the max-like magnitude (CMAG) all 
   the hypocentre data from 1964-83 have been 
   provided by the ISC. Please acknowledge the 
   ISC if this data is used in publications. 

2) Max-like values (CMAG) are in cols 55, 56 & 57
   and are given to 2 dec places. The method used
   to compute these values differs slightly from
   that described in our initial report (1) in
   several ways:

   (a) An alternative distance calibration cure is
       employed which enables data from the range
       delta = 30-180ø to be used (ref 2). This
       enables considerable amounts of PKP data
       to be included for many southern
       hemisphere regions. It also removes the
       problems associated with irregularities in
       the Gutenberg-Richter function. 

   (b) The noise levels used for the stations
       still vary with time but a uniform set of
       station terms are applied (listed in ref
       2). It is hoped that using an overall zero
       baseline for the total network 1964-83
       will reduce the problems associated with
       temporal changes in world network
       reporting.

   (c) Data loss due to clipping is a problem with
       larger events and to reduce the effects of
       this, magnitudes above mb 5.5 have been
       computed using a maximum-likelihood
       algorithm which does not make use of
       "absent readings" (see ref 3).

Ref (1)  Lilwall, R C & J M Neary, 1986.
         Redetermination of Earthquake Body-wave
         magnitudes (mb) using ISC data. AWRE
         Report 0 21/85, HMSO, London.
 
Ref (2)  Lilwall, R C, 1987. Empirical Amplitude-
         Distance/Depth Curves for Short-Period P
         waves in the Distance Range 20-180ø.
         AWRE Report 0 30/86 (Enclosed).

Ref (3)  Lilwall, R C, 1986. Some Simulation
         Studies on Seismic Magnitude Estimators.
         AWRE Report 0 22/86 (Enclosed).