the Genesis of Dalton's Atomic Theory. 157 



up in 1803, is otherwise in absolute contradiction with Dalton's 

 u brief historical sketch." 



Debus, however, refuses to agree in this conclusion, but 

 maintains that the date 1805 really represents Dalton's mean- 

 ing. The argument which he brings forward to explain 

 away the inconsistency which this involves is that the last 

 paragraph of the account quoted above from Dalton's " brief 

 historical sketch " is a disjointed note, quite distinct from the 

 rest, and does not bear any historical relation to the preceding 

 narrative. Making this assumption, Debus admits that Dalton 

 may have busied himself with the determination of atomic 

 weights and diameters at an earlier date, but supposes that he 

 did not conclude that the atoms must have different diameters 

 until 1805, and this in face of the fact that Dalton in 1803 

 drew up a table showing that the diameters of the particles 

 of all the gases he examined were different ! 



It is, however, clear from the MSS. (1810) that this con- 

 tention cannot be upheld ; the last paragraph follows in due 

 order upon the foregoing one, and its contents bear an obvious 

 relation to the preceding remarks. 



Another argument of Debus in favour of the later date is 

 that in 1805, in passing for the press a paper read in 1802, 

 in which reference is made to the theory of selective repulsion, 

 Dalton did not alter this, although he is known to have taken 

 advantage of this opportunity to make additions and alterations 

 in seA^eral papers published at the same period. This will 

 hardly surprise us when we remember that even as late as 

 1808 (N.S.p. 154) Dalton thought it worth while not only to 

 give a full account of the old theory, but even of the way in 

 which many objections to it might be met. 



Summing up the evidence as to the date at which Dalton 

 changed his view about the nature of diffusion, it is clear 

 that we must either admit the existence of a slip in the date 

 1805, written by Dalton in 1810, or else assume an absolute 

 discontinuity and lack of coherence in a statement of which 

 the continuity, both of sense and position, is obvious. Of 

 the two alternatives we still prefer the former as by far the 

 more reasonable. 



Debus, moreover, retaining the date 1805 for Dalton's 

 changed view about the diffusion of gases, further argues that 

 he retained his belief in the hypothesis M/S = C " more or less, 

 from 1801-1805 " (p. 362)/ Support for this idea is sought 

 in two or three passages contained in Dalton's note-book (1803) 

 and in Thomson's account of the atomic theory (' System of 

 Chemistry/ vol. iii. 1807). The first of these passages occurs 



