WHAT IS ASTRAGALUS HYPOGDOTTIS? 231 



A. epiglottis, which, however, is incompatible with either the 

 figure or the diagnosis. 



In the Species Plantarum ed. i. p. 759, the diagnosis of 

 Astragalus No. 1 of Hort. Cliff, reappears as a first synonym of 

 A. epiglottis, now described as " A. canlescens procumbens, legu- 

 minibas capitatis cordatis acatis reflexis, compUcatis . . . Habitat 

 in Hispania," with the fifth and sisth synonyms as quoted above. 

 This description and the Hort. Cliff, synonym can only apply to 

 A. epiglottis, the two other synonyms only apply to A.pentaglottis. 

 In Linnaeus's own interleaved copy in the possession of the 

 linnean Society, the word " r e fieri s " is struck out and replaced 

 by ''hirsutis" in his own hand. This alteration indicates thai 

 the l^umes of A. purpureus and not those of A. epiglottis or of 

 A. pentaglottis were now being aimed at. There is also an 

 addition of " Galloprovincia, Oriente" to the habitat. These 

 alterations were embodied in the second edition of the Species 

 Plantarum (1763), p. 1069, which, with this exception, reproduces 

 the exact words of the first. 



In Syst. Nat. x. p. 1171 (1759), xi. p. 1171 (1760), and xii. 

 p. 199 (not 199, as misprinted in the Mantissa), we find A. epi- 

 glottis with the diagnosis repeated exactly as in Sp. PL ed. ii., but 

 with all synonyms omitted. 



Although it was not tiU the Mantissa altera of 1771 that 

 Linnaeus separated the three species hypoglottis, pentaglottis, and 

 epiglottis, he had already become aware of the confusion, and 

 prepared the ground for the separation in his MS. notes in the 

 interleaved copy of ed. 2 of the Species Plantarum, also in the 

 possession of the Linnean Society. There, at p. 1069, the word 

 " hirsutis,'' which had crept into the diagnosis of the legumes of 

 A. epiglottis, is struck out again, thus cleaLring away an expression 



iung 



A. pentaglottis. To the synonyms is added " Glaux minima Riv. 

 tetr." The figure of " Glaux minima" in Rivinus, Icones Plant. 

 Fl. Irr^. Tetrapetalo,'^ is an excellent representation of A. epi- 

 glottis, whilst the lower figure of the same plate, " Glaux hispanica," 

 is an equally good one of A. pentaglottis. To the habitat are 

 added the words '' syhis montosis" but they are immediately 

 preceded by the erased words ^' sterilibus apricis," which, of 

 course, would have been the correct description of the haunts of 

 A. epiglottis. The erasion and substitution may have been a slip 

 of the pen, or may point to a confusion with the habitat of 

 A. purpureus. More important is the substitudon of the sign for 

 " perennial " instead of that for " annual." This is incorrect for 

 epiglottis, though correct for hypoglottis. So the confusion does 

 not seem to have been completely cleared up yet. On the inter- 

 leaf there is a fresh diagnosis of A. epiglottis, worded rather 

 differently from that of the Mantissa. It runs as foUows: — 



* The plates are not numbered. 



