As it has been clearly proved, however, (at least as I think) that this is entirely a false conception, and that, independent of any political or social institutions whatever, the greater part of mankind, from the fixed and unalterable laws of nature, must ever be subject to the evil temptations arising from want, besides other passions, it follows from Mr Godwin's definition of man that such impressions, and combinations of impressions, cannot be afloat in the world without generating a variety of bad men.

According to Mr Godwin's own conception of the formation of character, it is surely as improbable that under such circumstances all men will be virtuous as that sixes will come up a hundred times following upon the dice.

The great variety of combinations upon the dice in a repeated succession of throws appears to me not inaptly to represent the great variety of character that must necessarily exist in the world, supposing every individual to be formed what he is by that combination of impressions which he has received since his first existence.

And this comparison will, in some measure, shew the absurdity of supposing, that exceptions will ever become general rules; that extraordinary and unusual combinations will be frequent; or that the individual instances of great virtue which had appeared in all ages of the world will ever prevail universally.
I am aware that Mr Godwin might say that the comparison is in one respect inaccurate, that in the case of the dice, the preceding causes, or rather the chances respecting the preceding causes, were always the same, and that, therefore, I could have no good reason for supposing that a greater number of sixes would come up in the next hundred times of throwing than in the preceding same number of throws.

But, that man had in some sort a power of influencing those causes that formed character, and that every good and virtuous man that was produced, by the influence which he must necessarily have, rather increased the probability that another such virtuous character would be generated, whereas the coming up of sixes upon the dice once, would certainly not increase the probability of their coming up a second time.