http://network.nature.com/people/primatediaries/blog/2009/05/18/scientia-pro-publica-4-in-memory-of-stephen-jay-gould

The Greeks, being geometers, naturally thought that atoms differed among themselves in shape. Dalton, in whose time weight and measurements had grown important, maintained that the difference was one of weight, and he pioneered the concept of atomic weight.

For instance, 8 grams of oxygen will combine with 1 gram of hydrogen to form 9 grams of water. Suppose that water is formed through a combination of one atom of oxygen and one atom of hydrogen (the result would be a water molecule). In that case, one atom of oxygen would have to be eight times as massive as one atom of hydrogen. If hydrogen was supposed to have an atomic weight of 1, oxygen would have one of 8.

Of course, water might be made up of molecules containing any number of oxygen and hydrogen atoms. Dalton supposed it to be built up of one of each merely because that was the simplest possible solution. Until such time as more was known about molecular makeup, the values obtained for atomic weights would be dubious. In Dalton’s table of atomic weights (the first ever compiled), many were indeed wide of the mark.

Jumping ahead to 1811, under “Avogadro’s Hypothesis”:

It was clear that all gases expanded by the same amount as temperature rose, provided the pressure remained constant. In 1811 an Italian physicist, Amedeo Avogadro (1776–1856), suggested that this might mean that all gases — at the same volume, pressure, and temperature — were made up of the same number of particles. This came to be called Avogadro’s hypothesis.

If this is so, since water upon being broken up by an electric current decomposes into hydrogen and oxygen, with hydrogen having twice the volume of oxygen, then twice as many particles of hydrogen must be formed as of oxygen. This in turn makes it appear that water particles are not made up of one hydrogen atom and one oxygen atom, as Dalton had thought, but may be made up, at the simplest, of two hydrogen atoms and one oxygen atom. [...] Avogadro’s hypothesis, if fully applied, would explain a great deal about atomic weights and about the atomic constitution of compounds. Unfortunately, the hypothesis was largely ignored for half a century, and chemists remained unnecessarily confused in many ways during that time.

The neglect of Avogadro’s hypothesis stopped when Stanislao Cannizzaro promoted it at an 1860 conference in Karlsruhe. It looks like Faraday was writing during the period when atomic weights hadn’t yet been sorted out.

(I first read your blog when swansont linked to
levitation-and-diamagnetism-or-leave-earnshaw-alone)