cases of felo de se must have occurred, where, either from a lack of evidence or other reasons, the veil surrounding the event never having been fully lifted, some deaths which were suicidal may have been set down as accidental; the converse of this proposition being equally correct, as in cases of drowning or poisoning, for instance, deaths which may have been voluntary may easily have been ascribed to accident or error. With such contingencies, however, it is not my intention to deal, beyond their fleeting consideration, as a possible source of inexactitude, as we cannot go behind the records, and must accept the returns in the Coroner's dockets as final.

Contrary to expectations, this form of mortality does not follow what might seem to be the universal rule that the greater the population the greater the number of suicides. For the annual number of self-murders, as well as their ratio to a given figure of the population, is subject to marked variations, as the following figures show us plainly :

Looking at suicide, in its relations to color and race, it is somewhat humiliating for the white observer to be obliged to note that of the 636 recorded cases of felo de se only 8 were people of color, while, with the exception of one case, where the color was not entered on the docket, and is, therefore, unknown, the remaining 627 were white persons. A vast numerical but certainly most undesirable superiority for the intellectual white race! Creating a fact which will go far to strengthen the position of those who maintain that suicide is essentially a disease of the intellect and mind. Of these 8 cases, 6 were males and 2 were females; 4 of the former were married, 1 was single, and in 1 the condition in life was unknown. Of the females, 1 was married, and the other was single.

As in mortality from all causes in Philadelphia during the past decade, the number of male deaths annually occurring usually exceeded the number of female deaths; so also we have the male exceeding the female suicides in a proportion of nearly 5 to 1, there being 526 males to 110 female cases of self-murder, these taking place yearly in the following order:

Thus giving for the males a yearly average of 52.6 to an average of 11.0 for the weaker sex, with a general average of 63.6 yearly; the year 1875 showing the greatest number of suicides69 males to 17 females, or a total of 86-and the smallest number in any year occurring in 1878, when there were only 42 selfmurders, 35 of which were males, and 7 females.

The rule observed by mortality tables in general, that the deaths of adults are usually in excess of those of minors, is also observed in the mortality of suicide, there being only 24 instances of suicide by minors among the 636 cases with which we are occupied. Of these, 14 were male, and 10 were females.

Considering self-slaughter in relations to age, we are not surprised to observe, that appearing about the age of puberty and steadily increasing, it attains its maximum when the illusions of youth having been dispelled by the stern realities of life, the struggle for existence is keenest; afterwards, gradually declining in numbers as the health begins to tail, as the bodily vitality is exhausted, and the interest which individuals manifest in worldly events becomes more and more of a retrospective character.

We find, therefore, that, according to age, self-murder occurred in the following order of frequency:

The highest number at any period occurring between the ages of 30 to 40 years, the lowest number being between the ages of 80 to 90 years; the female following the male sex in its successive stages of increase and decrease, except that between the ages of 20 to 30, the weaker seems to attain its maximum more rapidly than the stronger sex; who, however, largely exceed the former, save in the two earlier periods, where the proportion of males is only 13 to 1 and 3 to 1, in excess of the females.

In opposition to the experience of European observers, who state that self-destruction occurs more frequently among the unmarried of both sexes, than amongst the married, we find that in this centre at least, and for the period with which we are dealing, such is not the case. Whether it be because in Philadelphia, as exhibited by statistics, the most popular time for marriage for both sexes is between the ages of twenty to twenty-five years; while as we have just seen, suicides are not of frequent recurrence until a more advanced period of life has been reached, or whether it be due to other causes, only a profound study of observations extending over a more lengthy period of time, and comprising larger masses of population, can teach us. But, out of the 636 cases of suicide with which we are now concerned, 444 were married, 138 were single, and in 54 persons the condition in life was not determined; while, according to sub-division by sexes of the 444 married individuals, 370 were men, 74 were women; and of the unmarried people 109 were males, and 29 were females. Of the 54 whose

condition in life was not recorded, 47 were of the male and 7 of the female sex. And we must greatly regret that no record has been preserved as to how many of those classified as married were widowers or widows, whether they were blessed with children, or childless; as also whether any of those recorded as unmarried were living in a state of concubinage, or otherwise.

Though, owing to the vast number of observations which have been made, it is a well-known fact that, in the death-rate from ordinary causes, meteorological phenomena appear to exert considerable influence in determining the frequency and amount of mortality by months and seasons, we are unfortunately as yet in ignorance as to the role played by such phenomena, as determining factors in the more or less frequent occurrence of suicide; our observations being neither accurate nor complete enough to enable us to deduct conclusions of any real value. Nevertheless, through the kindness of the Chief Signal Officer of the Army, I have been enabled to obtain a summary of the meteorological observations made at the United States Signal Station in the city of Philadelphia, during the past nine years.* These observations have been arranged by groups of months, and the following table gives the means and totals, together with the number of suicides for each group of months:

* Prior to 1873, there was no Signal Station in Philadelphia.

According to which we find the real or apparent exceptions so numerous that no satisfactory conclusion can be arrived at. The influence of barometric variations alone seems negative; as do also the thermometric changes, while the amount of rain or snow fall, and the number of days on which it fell, give results entirely at variance with the number of suicides presented. The prevailing direction of the wind alone would seem to show some connection with the large number of suicides occurring in certain groups of months. Considering, however, together with all the data presented by the different observations in the table, they would seem―to a certain extent at least—to indicate that a low barometric pressure, accompanied by a high thermometric registry with sudden fluctuations from a low to a high temperature, together with much moisture and prevailing south-west winds, might somewhat account for the frequency of self-murder in the spring and summer months.

Again, we find that in these observations the deaths from suicide, as occurring by months, differ somewhat from the results published by the Health Office, upon the successive monthly order of occurrence of deaths from all causes in Philadelphia, during a period of fifteen years,-1861 to 1875.