The returns received were sent to the Massachusetts Bureau of Statistics of Labor, where, by the courtesy of the chief of the bureau, they were collated during the spring and summer of 1890, under the special direction of the chief clerk, in accordance with a previously arranged scheme of tables. The general plan of arrangement adopted was to class the schedules with reference to employers, first alphabetically by states and towns, and second alphabetically by population. The schedules were then classed with reference to employees, first by men and women, and second by place of birth. The various statistical devices used in the Massachusetts Bureau were employed in tabulating the material, and greatly facilitated the work. Fifty large tables were thus prepared, and by various combinations numerous smaller ones were made. The classification thus adopted made it possible to give all the results either in a general form or with special reference to men and women employees, the native born and the foreign born, and to all of the branches of the service. It was also possible to study the conditions of the service geographically, and with reference to the population and to other industrial situations. The most detailed tables made out concerned the wage question, including a presentation of classified wages, average wages with the percentage of employees receiving the same wages as the average, and also more or less than the average, a comparison of wages paid at different times and of wages received in domestic service and in other employments. For the purposes of comparison, the writer also classified the salaries paid to about six thousand teachers in the public schools in sixteen representative cities, as indicated by the reports of city superintendents for the year during which information concerning domestic service had been given on the schedules. Through the courtesy of a large employment bureau in Boston, the wages received by nearly three thousand employees were ascertained and used for comparison. The most valuable results of the investigation possibly were those growing out of the consensus of opinion obtained from employers and employees regarding the nature of the service considered as an occupation. The greater proportion of these tables can be found in Chapter V.