Chart I has been prepared from the above table and shows at a glance the relative prevalence of lynching during the twenty-two years, 1882–1903.[212] The solid line, representing the number of persons lynched, begins at 114 in 1882, goes up to 211 in 1884, falls to 122 in 1887, rises again to 176 in 1889, falls again the following year to 128, and then rises to 235 in the year 1892. From 1892 on the line shows a general downward tendency. The large number in 1884 was due to the work of vigilantes in Montana and Colorado. According to the record, 40 horse thieves and cattle thieves received summary justice in Montana during the year and the lynching of seventeen cattle thieves was reported from Colorado. The large number in 1892 was due to the lynching of negroes in the South. These facts are shown in another form in Chart IV. For purposes of comparison the broken line has been drawn on Chart I to represent the number of persons legally executed during the twenty-two years. The figures for legal executions have also been taken from the Tribune record, but no further verification has been made of them than merely to foot up the lists of names and verify the totals. The Tribune record of legal executions is no doubt incomplete, but it is here given for what it is worth. In a general way the broken line follows the solid line with the exception of the year 1892 and the last five years. In the review of the year 1881, the Tribune reported ninety-one legal executions but gave no record of lynchings. To indicate the direction of the broken line previous to 1882 a short line has been drawn at the left of the chart. On the whole, Chart I seems to indicate a relationship between legal executions and lynchings. There is an indication that the upward tendency in the broken line since 1887 may have contributed to the downward tendency in the solid line since 1892—the punishment of crimes by law may tend to make recourse to lynching less frequent. Another factor is needed, however, to make this reasoning conclusive; some way of measuring the amount of crime in the country during these several years is requisite. For this purpose the writer took the number of murders reported by the Tribune and plotted them on a trial chart in connection with the number lynched and the number legally executed. The result was so unsatisfactory and inconclusive, however, that any attempt to make such a comparison was abandoned.[213]