I began to get interested. Supposing I dug out _all_ the homicide cases I had ever tried, what would the result show as to motive for the killing? Would drink and women account for seventy-five per cent? Mentally I ran my eye back over nearly ten years. What _other_ motives had the defendants at the bar had? There was Laudiero—an Italian “Camorrista”—he had killed simply for the distinction it gave him among his countrymen and the satisfaction he felt at being known as a “bad” man—a “capo maestra.” There was Joseph Ferrone—pure jealousy again. Hendry—animal hate intensified by drink. Yoscow—a deliberate murder, planned in advance by several of a gang, to get rid of a young bully who had made himself generally unpleasant. There was Childs, who had killed, as he claimed, in self-defence because he was set upon and assaulted by rival runners from another seaman’s boarding house. Really it began to look as if men killed for a lot of reasons. I wanted to call up my friend and ask what kind of killings counted. Did he simply want to know why men murdered one another? He couldn’t possibly mean that I was to attempt to explain why they saw fit to exterminate each other by means of capital punishment? Or ran over one another in trains and automobiles? Or allowed each other to die from unsanitary conditions? Or lynched one another?—there was only _one_ reason for that I knew. Or killed themselves? Nor did he mean to have me go into the question of why they killed elsewhere—in Naples, Sicily, Constantinople, and so on. No; what he wanted to find out was why men in the United States of America killed other men of their own kind without malice aforethought—legal and quasi-legal killings excluded. Moreover, he wanted to know from the actual personal experience of those who had weighed the evidence as to their motives in a sufficiently large number of cases to be representative. One consideration at once suggested itself. How about the killings where the murderer is never caught? The prisoners _tried_ for murder are only a mere fraction of those who _commit_ murder. True, and the more deliberate the murder, the greater, unfortunately, the chance of the villain getting away. Still, in cases merely of suspected murder, or in cases where no evidence is taken, it would be manifestly unfair arbitrarily to assign motives for the deed, if deed it was. No, one must start with the assumption, sufficiently accurate under all the circumstances, that the killings in which the killer is caught are fairly representative of killings as a whole. All crimes naturally tend to divide themselves into two classes—crimes against property and crimes against the person, each class having an entirely different assortment of reasons for their commission.