378 Notices respeciinq New Books. 



therefore, that there should be no other proportions of combina- 

 tion between these two elements, unless indeed it be one that is 

 expressed by an even sub-division of one of tliese proportions^ as 

 for example, 5 A to 7, H'i, 14, &:c. of B; in which case the 

 10^ being resolvaljle into three portions of i^ each, the number 

 expressing the relative weight of an atom of B must be reduced 

 to 3-2- instead of 7, and consequently the several proportions of 

 7, 10', 14, and 21 of B will be resolved respectively into 2, 3, 

 4, and C atoms of B. 



" To verify the numbers expressing the relative weights of an 

 atom of A and B, (supposing that of A to be assumed as 5, and 

 that of B as 7,) let them each be examined in their separate 

 compounds with a third liodv C. Then, suppose that iu the 

 simplest binary compound of A and C, analysis discovers 3 parts 

 by weight of C, to 5 of A, it is assumed, that, as 5 is taken as 

 tlic numerical expression of an atom of A.., the number express- 

 ing an atom of C must be 3 ; and cotiseqnenibi, if this mode of 

 estimation be just, it will Tilso be found by experiment that in the 

 simplest combination of C with B, 7 parts of B will unite ex- 

 r.-:tly with 3 parts of C. This supposes indeed that this simple 

 combination of one atom of each body is known by experiment ; 

 but even if this should not be the case, the general principle will 

 not be contravened, if, instead of a single portion of C being found, 

 there sliould be a double, triple, or quadruple portion, provided 

 the radical number or common divisor is 3, that of B being 7. 



" To illustrate this by an example from Mr. Daltoa (in which 

 however the numbers assumed are not perfectly accurate though 

 sufficient for the present purpose). The substance of which as 

 far as we yet know the smallest relative weight enters into che- 

 mical combination is hydrogen, and on this account the weight 

 of its atom is assumed as unity, and is the standard of compa- 

 rison for the relative weight of the atom of all other bodies. The 

 only compound of hydrogen and oxygen that we know is water, 

 in which tl'.e oxvgen is to the hydrogen as 7 to 1. The number 

 7 therefore is assumed as the rehitive weight of the atom of oxy- 

 gen, and water is a. binary compound containing an atom of hy- 

 drogen with an atom of oxygen iu every atom of water. Sul- 

 phuretted hydrogen is composed, according to Mr. Dalton, of 

 13 parts by weight of sulphur, and 1 of hydrogen. If it be as- 

 sumed that an atom of sulphuretted hydrogen contains an atom 

 of sulphur united to an atom of hydrogen, the relative weight of 

 an atom of sulphur unist be 13. To prove that this number 13 

 may be assumed as the weight of an atom of sulphur, let it be 

 examined in its compounds with oxygen, and, if correct, all the 

 con^pounds of these two elements will contain for every 13 parts 

 by weight of sulphur, either 7j or 14, or 21, &c. parts of oxy- 

 gen. 



