But though these parts be alike in their nature, they are very different in their quantity and number; and this difference must appear in the effect as well as the similarity.

Now as the view they present is in both cases full and entire, and comprehends the object in all its parts, it is impossible that in this particular there can be any difference; nor is there any thing but a superior vivacity in the probability, arising from the concurrence of a superior number of views, which can distinguish these effects.
Here is almost the same argument in a different light.

All our reasonings concerning the probability of causes are founded on the transferring of past to future.

The transferring of any past experiment to the future is sufficient to give us a view of the object; whether that experiment be single or combined with others of the same kind; whether it be entire, or opposed by others of a contrary kind.

Suppose, then, it acquires both these qualities of combination and opposition, it loses not upon that account its former power of presenting a view of the object, but only concurs with and opposes other experiments, that have a like influence.