or infected water supplies, but were continued at intervals from the middle of July to October, a few being reported in each week.

Professor Sedgwick says: "The epidemic was of a slow, insidious character, as if slowly passing from person to person, and from house to house; and as has been stated, it was confined to one, and that an unsanitary section of the city."

It was recommended that an agent of the Board of Health should be detailed to give special attention to disinfection among the houses in this district where the fever existed.

Limited outbreaks were also reported as having occurred at the Mount Hermon School in Northfield, at the State Prison, and at Rowley, Mass., but the total number of such outbreaks has been much less than it has in previous years.

Five men were taken ill at the State Prison in October, and the secretary visited the prison and inspected its sanitary condition. The general condition of the buildings, cells, kitchen, hospital and other apartments appeared to be good. A sewer was being constructed in the grounds outside, upon which the convicts were at work, but none of those who were taken ill had worked upon this

sewer.

The water supply is the water of Mystic Lake in Medford, occasionally alternating with water from the general, or Cochituate and Sudbury supply. The Mystic supplies the cities of Somerville, Everett and Chelsea. An inquiry of the health authorities of these cities showed that typhoid fever was not unusually prevalent among them at the time of the occurrence of the cases at the prison.

Attention was then turned to the milk supply, which was obtained from a contractor doing business in Charlestown, having a large business, and obtaining his milk from about one thousand dairies in Maine and southern New Hampshire. The milk came in each morning over the Western Division of the Boston & Maine Railroad, being collected at a station in southern New Hampshire.

The conditions prevailing at the milk distributing station at Charlestown were fairly good, and all the men connected with the work of distribution and present at the station appeared to be in good health and free from infection.

One man who had charge of the stable, opening out of the distributing room, was given a vacation September 8 and went away, but was soon after attacked with typhoid fever and went to hospital, where he was at the time of the inspection (October 22). This man

had no connection with the work of distributing or handling of the milk, and his illness was undoubtedly in the relation of effect and not of cause, since he and his family used the milk furnished by this contractor.

The supply furnished to the prison was forty cans four times a week, and these were not always from the same dairies, but so far as could be learned were from several dairies along the line of a railway extending from southern New Hampshire to Maine. It was deemed impracticable to continue the investigation in this direction.

Another limited epidemic, in which the total number of cases was only 7, was that of the Mount Hermon School, on the west bank of the Connecticut River in Northfield. An inspection of the school was made and advice given relative to its water supply and sewage disposal.

In the reports of the Board from 1890 to 1895, inclusive, a table has been published annually presenting the deaths and death-rates from typhoid fever for a series of years in the cities of the State, these cities being arranged in the order of their population. In the present report, page xiv, this table has been rearranged, the statistics relating to typhoid fever being presented in five-year groups, and in each period the cities are arranged in accordance with the intensity of their death-rates from typhoid fever. North Adams, which became a city in 1895, has been added to the list, making a total of thirty-two cities.

By means of this table the comparative rank of the different cities in each period and the rise or fall in the typhoid death-rate from one period to another are shown.

While the general improvement in the death-rate from typhoid fever, coincident with the introduction of better water supplies, is shown in the latter part of the report in the portion relating to vital statistics, it also appears in this table, where the conditions of certain cities are considered separately.

In some instances radical changes have taken place in the rank of cities, comparing early with late periods. Boston has maintained a position quite uniformly intermediate in the list, and usually slightly below that of the State. It would undoubtedly have been still lower were it not for its intimate connection with the surrounding territory and the presence of many hospitals, in which patients from other cities and towns are admitted for treatment. The manufacturing cities, Holyoke, Springfield, Chicopee, Lowell, Lawrence, Fall River and North Adams, have generally had high death-rates

throughout most of the time included in the table. On the other hand, the smaller residential cities, Waltham, Newton, Medford, Everett and Woburn, have usually been found near the foot of the list, with low typhoid death-rates.

Holyoke, which in the first three five-year periods was at the head of the list of high death-rates, has shown a decided improvement in the fourth and fifth periods.

Pittsfield, also, from a position above the mean, has dropped below the mean, and in the fifth period had the lowest death-rate from typhoid fever among the cities.

On the other hand, Newburyport has changed its position from a generally low rank to a position near the top of the list.

The death-rates in this table are annual death-rates per 10,000 of the living population. The population had increased, during the whole period indicated in the table, about 83 per cent. (comparing the mean population of the first five years with that of the last five).

1886-1890.

1871-1875.

1876-1880.

1881-1885.

1891-1895.

This comparison may be carried still further by an examination of the following table for the year 1896, which has been made up from the returns required by chapter 218 of the Acts of 1894. The figures in this table may differ slightly from those which are obtained from the Registration Report at a later date, but the difference is usually very small. It may also be noted that the results of a single year are not so conclusive as those of longer periods, and those of small cities are less conclusive than those of large ones, according to the usual principle of the limits of error.

Haverhill is found at the top of the list in this table for 1896. With regard to the high typhoid death-rate of Haverhill in 1896 reference may be made to Professor Sedgwick's report to the board of health of that city, quoted near the close of this volume, under the head of Health of Towns."

Lynn in the last decade, 1886-95, maintained a record for typhoid fever mortality better than the average. In 1896 the death-rate of Lynn from typhoid fever is only the sixth from the top of the list.