684 Barnhart : Dates of Elliott's Sketch 



win called Elliott's attention to the fact that he had named it D. 

 latifolia in previous letters, so this name was adopted in the new 



issue. 



mu 



:h for the first number. The date of the second number 

 cannot be decided quite as accurately, but we are safe in saying 

 that it was within a very few days of January i, 1 8 1 7, and probably 

 before that date. I have seen no original copy of its cover, but there 

 is no doubt that it bore the date 1 8 16. De Candolle (Syst. 2 : 703. 

 1 821), doubtless writing with the first four parts before him in their 

 original covers, says : "Elliot (Steph.). A Sketch of the Botany 

 of South-Carolina and Georgia. In-8. Charlestown Fasc. 1, 11, 

 18 16, 111, iv, 1817/' Gray, in the note quoted at the beginning 

 of this paper, says "before January, 18 17," and I believe he is 

 correct. Alterations suggested in Baldwin's letter of Nov. 1 1 

 were made use of in the correction of No. 1, the reissue of which 

 occurred with No. 2, so they could not easily have appeared be- 

 fore December. Baldwin, writing to Darlington during the first 

 half of February, 181 7 (Reliq. Baldw. 202), says that he has not 

 yet received them, u although noticed in Charleston several weeks 



ago ; " 



J 



1 8 17 ; and it seems altogether probable that they were published 

 in the latter part of December, 1816. At all events, they were 

 almost certainly dated 18 16, and in such cases the burden of proof 

 always rests with those who would claim that the date as printed 



is incorrect. 



It may be as well here as anywhere to speak of the pagination 

 of the different parts, for here Gray was far astray in his guesses. 

 In his first note he proves (to his own satisfaction) that No. 4 

 included p. 466 ; while his second note shows clearly that No. 5 

 began with p. 401; but he calls no attention to the discrepancy, 

 nor does he offer any new suggestions. As a matter of fact, each 

 number consisted of as nearly one hundred pages as was possible 

 without breaking a signature. Each signature consisted of eight 

 pages, so that one hundred pages would make twelve and a half 

 signatures. To avoid this, the first part consisted of twelve sig- 

 natures (96 pages), the second of thirteen (104 pages), the third 

 of twelve, and so. on, to the last of volume one. The evidence I 

 have to offer upon this point is incontrovertible. We have seen 



