Where the Peters volumes covered Neotropical birds studied by Hellmayr, he was accepted as a reliable source. As a result, Peters omitted names synonymised by Hellmayr, even if they post-dated Sharpe's 'Handlist', contrary to Peters's policy for other regions. However, for livr.5 Peters himself was more cautious than Hellmayr and noted the ref. as late 1830 or early 1831, but this may have been because Hellmayr dated livr.5 names as either 1830 or 1831, with 1831 apparently preferred later. Mathews himself accepted livr.5 as 1830 and used this later (e.g. in his 'Systema Avium Australasianarum', 1927-1930).

This confusion is reflected in how livr.5 names have been cited in Peters volumes. I have a copy of the Lesson publication and went through livr.5, i.e. pp.321-400, to sample how they were handled in Peters, with the following result (an asterisk means only the page mentioned, i.e. no indication of 'livr.5'):

	p.323*  Fregilupus  		15:103 - 1831
	p.324   Corcorax  		15:160 - 1830
	p.329   Strepera  		15:170 - 1830 or 1831
	p.341*  Temnurus  		15:250 - end 1830 or early 1831
	p.347   Batara  		7:155 - between Sept. 1830 and Feb. 1831
	p.349   Graucalus Macei  	9:169 - 1831
	p.360*  Attila  		8:186 - 1830
	p.363*  Querula minor  		8:239 - 1830
	p.372*  Corvinella  		9:341 - 1831
	p.373*  Tchagra  		9:320 - 1831
	p.374*  Schetba  		9:365 - 1830
	p.375   Taraba  		7:157 - between Sept. 1830 and Feb. 1831
	p.375*  Pitohui  		12:45 - 1830
	p.376   Lanius vestitus  	7:167 - between 1830 and Feb. 1831
	p.384*  Todirostrum  		8:87 - 1831
	p.385   Bias  			11:376 -- ?1830
	p.385   Muscylva  		11:472 - ?1830
	p.386   Tchitrea  		11:479 - ?1830
	p.387*  Yetapa  		8:182 - 1831
	p.387*  Muscipipra  		8:186 - 1831
	p.387   Arses  			11:514 - ?1830
	p.389   Miro  			11:562 - ?1830
	p.392   Muscicapa Delalandi  	7:277 - late 1830 or early 1831
	p.397   Merulaxis  		7:281 - end of 1830 or beginning of 1831
	p.397   Merulaxis ater  	7:281 - end of 1830 or beginning of 1831
	

All things considered, I accept the dates of publication of the livraisons as established by Mathews in 1925 (as quoted from the Zoonomen website). As indicated above, the confusion can be traced through Peters to Zimmer and Hellmayr. Therefore my conclusion is for HBW to accept 1830 as the date for all names cited from livr.5. Thus 1831 only applies to names published after p.400."

Edward's letter (2004.11.22) is as follows:

	The Bibliographie de France was a weekly publication produced by the printing trade. 
	Within each weekly issue the bulk of the content concerns new French books 
	(including livraisons of part works); there are other shorter sections that include 
	English and German books, but I spent no time looking at those, 
	and there are further sections dealing with certain kinds of other print. 
	The normal issue was 16 or 24 pp.  

	The information relative to Lesson's Traité, that was obviously found by Mathews (and Sherborn) 
	can be documented clearly for others to find and is as follows:
 
In 1830:
 
    in issue 7 dated 13 Feb. [pp. 97-112] on p.107: No.910 records the first livraison, 
    made up of 5 sheets of octavo [thus 80 pp.]
 
    in issue 19 dated 8 May [pp. 289-304] on p. 298: No.2538 records the second livraiosn, 
    made up of 5 sheets of octavo [thus 80 pp] plus 15 plates, 
    and a further 15 plates from the first livraison.
 
    in issue 28 dated 10 July [pp. 457-480] on p. 473: No.3932 records the third livraison, 
    made up of 5 sheets of octavo [thus 80 pp] plus 15 plates
 
    in issue 39 dated 25 Sept. [pp. 633-648] on p. 644: No.5177 records the fourth livraison, 
    5 sheets of octavo [thus 80 pp] plus 15 plates
 
    The index for the year lists only these four entries for this publication.
 
    I checked every issue from No.40 to the year end to see if the Traité appeared 
    (without having made the index); it did not. 
    This is not hard to check as the titles are arranged roughly in alphabetical order 
    in section I of each issue.
 
In 1831:
 
   in issue 10 dated 6 March [pp. 137-152] on p. 147: No.1100 records the sixth livraison, 
   5 sheets of octavo [thus 80 pp] plus 15 plates
 
   in issue 15 dated 9 April [pp. 201-216] on p. 213: No.1635 records the seventh livraison, 
   5 sheets of octavo [thus 80 pp] plus 15 plates
 
   in issue 24 dated 11 June [pp. 337-352] on p. 347: No.2732 records the eighth livraison, 
   5 sheets of octavo [thus 80 pp] plus 15 plates
 
  The index for the year records just these three entries for this publication.
 
  I checked every issue from No.1 to No.9 to see if the Traité appeared 
  (without having made the index); it did not.
 
  I did not search the 1832 volume for the last part. 
  The idea that there is a 'last' part is based on the fact that 8 x 80 = 640 
  and there are 659 pages in the finished work 
  (not counting the introductory pages to the text volume and to the separate volume of plates). 
  Note however that Lowe (1925, Ibis, p.483) thought, or assumed, that p. 650 was part of 
  livraison 8 (and thus implied that this livraison was not limited to 80 pages, 
  although I suspect he probably didn't even count). 
  Zimmer noted this and also that the plate count total seemed to exceed by one the number of 
  plates issued.
 
  Note that Zimmer thought (and perhaps Mathews said) that these dates were 'dates of receipt'; 
  they are not, they are the dates of the weekly publication. They may or may not be reporting 
  a mandatory donation during the previous week to some central library (as happens in the UK), 
  but they are certainly not dates of receipt.

  [APP note: Mathews does appear to regard them as dates of receipt, his wording
  on p.80 of the Bibliography of The Birds of Australia is "This was published in
  eight livraisons, the receipt and contents of which are acknowledged in the Bibl. de France as
  follow: ..."]
 
  I also examined the text of Mathews (1925) and I conclude that Mathews, 
  who refers ONLY to the Bibliographie de France had found nothing new and 
  was still guessing as to the date of the fifth livraison. 
  Nor is it obvious from the cadence of publication that November 1830 would be right. 
  Issue 7 was only 5 weeks behind issue 6; had that, in turn, been 5 weeks earlier it 
  would have appeared in the last week of January. Actually it would seem that Mathews 
  did not even find the data from issue 10 in 1831 as he was still unsure about livraison 6.
 
  In the circumstances I revert to my previous view that due to lack of proof for 1830 
  Livraison 5 must be treated as 1831.