>-■ 



i 



1941] MERRILL, DATES OF PUBLICATION OF WIGHT'S ICONES 223 



F 



The record is complete except for volume one, number one (plates 

 l_2o/, volume two (parts 1-4, plates 319-736), and volume three (parts 

 1_3^ plates 737-1046). A fascicle cover for volume 1, no. 1, is dated 

 1838, and in his prospectus Wight states that no. 1 was to appear in July 

 of that year; thus 1838 can safely be accepted as the date of issue of 



this part. 



It is worthy of note that the first twenty plates forming number one 



bear the numbers 35, 73, 160, 171, 176, 178, 189, 197, 198, 198[bis]; 



198[ter], 205, 334, 339, 341, 342, 396, 467, 513, and 723. These figures 



apply to the numbers of the species as described in Wight and Arnott's 



'Trodromus" published in 1834, These plates, in sequence, should be 



numbered 1 to 20, for while Wight added the forms 3\^, ^\, jf^, 

 etc., in his published explanations of the plates and brief descriptions, 



he did not commence to print these fractional forms on the plates until 



number two appeared. This system of numbering is obscure until one 



realizes that where the fractional form of numbering appears on the 



plates and in the text, the numerator represents the sequence of plate 



numbers for the "Icones" (this to be cited), and the denominator is 



the species number in Wight and Arnott's 'Trodromus florae Indiae 



Orientalis'' (1834), 



After number one appeared, Nathaniel Wallich, then Superintendent 

 of the Botanic Garden at Calcutta, placed at Wight's disposal the 

 unpublished drawings prepared under the direction of William Roxburgh, 

 and now preserved in the library of that institution. The use of these 

 Roxburgh drawings is explained by Wight in a notice issued with the 

 second number of the "Icones"; these illustrations, as reproduced by 

 Wight, all bear the legend '^Roxburghianae.'' This is mentioned to 

 emphasize the importance of these particular illustrations, for they 

 graphically represent numerous species actually described by Roxburgh 

 in his ^Tlora Indica" ed. 1, 1(1820), 2(1824), and ed. 2, 1-3(1832). 



Wight's original plan of publication was to issue twenty plates, with 

 descriptive text, as individual numbers, and this plan was approximately 

 followed for those parts forming volume one. " Beginning with volume 



two this plan was changed to one involving fewer and larger parts, the 

 four forming volume two containing from 98 to 116 plates each, this 

 system being continued to the end of the work. In reference to volume 

 two, I have no information as to the dates of issue other than Wight's 

 own statement (Icones 6: viii. 1853) that of volume two, plates 319-736 

 were issued from 1840 to 1843; the title page date is 1843. As plates 

 319-416 form part one, we can rather safely assume that this part was 

 issued in 1840. Part four includes plates 632 to 736 and clearly appeared 

 in 1843, This leaves plates 417 to 631, forming parts two and three, for 



