party was split into the Stalwart and Half-Breed factions. Accordingly
neither party had one dominant boss, or one dominant machine, each being
controlled by jarring and warring bosses and machines. The corruption
was not what it had been in the days of Tweed, when outside individuals
controlled the legislators like puppets. Nor was there any such
centralization of the boss system as occurred later. Many of the members
were under the control of local bosses or local machines. But the
corrupt work was usually done through the members directly.

Of course I never had anything in the nature of legal proof of
corruption, and the figures I am about to give are merely approximate.
But three years' experience convinced me, in the first place, that there
were a great many thoroughly corrupt men in the Legislature, perhaps a
third of the whole number; and, in the next place, that the honest men
outnumbered the corrupt men, and that, if it were ever possible to get
an issue of right and wrong put vividly and unmistakably before them
in a way that would arrest their attention and that would arrest the
attention of their constituents, we could count on the triumph of the
right. The trouble was that in most cases the issue was confused. To
read some kinds of literature one would come to the conclusion that the
only corruption in legislative circles was in the form of bribery by
corporations, and that the line was sharp between the honest man who was
always voting against corporations and the dishonest man who was always
bribed to vote for them. My experience was the direct contrary of
this. For every one bill introduced (not passed) corruptly to favor a
corporation, there were at least ten introduced (not passed, and in this
case not intended to be passed) to blackmail corporations. The majority
of the corrupt members would be found voting for the blackmailing bills
if they were not paid, and would also be found voting in the interests
of the corporation if they were paid. The blackmailing, or, as they were