362 THE AMERICAN JOURNAL OF SOCIOLOGY
affected by revivals so far as these are accessible to the student. Those who come into the various local churches by transfer from other churches are for obvious reasons left out of the accounting.
The statistics of the total number of members in the Metho- dist Episcopal Church of the entire country seem to have been fairly well kept from its first General Conference in this coun- try in 1773. But the statistics of probationers and of addi- tions are less satisfactory for statistical purposes. For this reason the figures given from the Methodists are for gains and losses in members as officially reported, except for the revival of 1877 where those for probationers are used. The Congre- gationalists of Connecticut began the collection of statistics as early as 181 6. But the earliest figures were meager and for members only and cannot be trusted for the purpose of this paper. A little before 1830 figures showing original additions begin to appear in that and other states. From 1830 the sta- tistics of additions on "profession" or "confession of faith" in the Congregational churches of New England and of "bap- tisms" in the Baptist churches are increasingly full. Statistics of this class, drawn from the records of the local churches annually by the clerks or sometimes by the pastors, are by far the most trustworthy of ecclesiastical statistics that we have for the period they cover. Their defects are of comparatively little importance, errors in the long run balancing each other.
The value of the statistics of these additions to churches lies in the fact that they are usually accepted as, and are on the whole the best test of the results of revivals. They are the most used by the advocates of revivals. Of course they are but one of several tests, most of which will come within the scope of those other studies which have been pointed out as needed.
A brief notice of the few figures for earlier revivals in New England may be properly taken at this point. In the valuable Contribution to the Ecclesiastical History of Essex County, Massachusetts, statistics are given from the original church records of the additions, original and by letter, the latter in those times being few in number, in one total, from 1701 down
