To be sure, this is theory. The practical question is, not how would the law operate, but how did it operate? This brings us again into the dangerous field of statistics. Now, if we may suppose that the witch cases known to us are a safe basis of comparison, the reign of James, as has already been intimated, shows a notable increase in witch executions over that of Elizabeth. We have records of between forty and fifty people who suffered for the crime during the reign of James, all but one of them within the first fifteen years. It will be seen that the average per year is nearly double that of the executions known to us in the first part of Elizabeth's rule, and of course several times that of those known in the last part. This increased number we are at once inclined to assign to the direct and indirect influence of the new king. But it may very fairly be asked whether the new statute passed at the king's suggestion had not been in part responsible for the increased number. This question can be answered from an examination of those cases where we have the charges given. Of thirty-seven such cases in the reign of James I, where the capital sentence was given, seventeen were on indictments for witchcrafts that had not caused death. In the other twenty cases, the accused were charged with murder.[16] This means that over two-fifths of those who are known to have been convicted under the new law would have escaped death under the Elizabethan statute. With all due allowance for the incompleteness of our statistics, it seems certain that the new law had added very considerably to the number of capital sentences. Subtract the seventeen death sentences for crimes of witchcraft that were not murder from the total number of such sentences, and we have figures not so different from those of Elizabeth's reign.