'224: THE JOURNAL OF BOTANY 



That he, not having pubUshed it there, should not allude to it in 

 any of his later \Yorks, is less surprising, for it must be remembered 

 that when he left Holland he was separated from Cliffort's her- 

 barium, and possibly — indeed probably — never saw this specimen 

 again. There are other instances of his having at the date of the 

 Species Plantarum forgotten the contents of Cliffort's garden and 

 herbarium in cases where those contents had not been published 

 in the volume of WiQHortus CUjfortianus. For example, Cliffort's 

 herbarium contains a fine specimen of Trionfetti's " Carduus 

 pycnopolyceplialus palustris " with the plumose pappus of a 

 Cirsium, whereas the specimen in Herb. Linn, of Carduus polyan- 

 tJiemus, to which Trionfetti's plant is referred in the Mantissa, is a 

 true Carduus with simple pappus, grown at Upsala, and is in fact 

 C. multiflorus Gaud. 



The remaining five specimens are all in Linnteus's herbarium, 

 and are all marked " epiglottis " in his own hand. 



(3) From Gouan, and so marked by LinnEeus himself. It is 

 certainly A. pentaglottis, and the name has been altered to peiita- 

 glottis by Sir J. E. Smith. The unmistakable legumes are well 

 represented. 



(4) A specimen marked by Linnaeus "A. 107. C." The meaning 

 of this indication is not known. Dr. Daydon Jackson, who has 

 recently discovered the clue to several of these Linnean marks, is 

 unable to explain it. The sheet to which it is attached is pinned 

 to an upper sheet bearing a piece of A. Glaux sent from Spain by 

 Loefling. Although marked ''epiglottis" it is certainly Glaux, 

 like the plant on the upper sheet. 



(5) Absolutely typical epiglottis grown in Hort. Upsalensis. 

 Smith as well as Linnaeus has duly marked this " epiglottis.'' I 

 suspect that this specimen was added to the herbarium after the 

 publication of the second edition of the Species Plantarum, and it 

 may have been the warning which caused Linn^us to correct in 

 the Mantissa his account of A. epiglottis. 



(6) A specimen on a sheet pinned underneath that of the next 

 specimen (no. 7). It is so poor a specimen that I dare not 

 determine it. There is a sessile head of two or three largish 

 upright glabrous legumes, rather hke those of A. bcsticus. At any 

 rate, it has nothing to do with the species we are considering. 



(7) The true type of A. liypoglottis. It is a plant sent by 

 Gerard, marked (by him) N. 28, and named ''epiglottis" in 

 Linnaeus's handwriting, but corrected in pencil by Smith as 

 follows: " Hypoglottis, vide descr. optimam in Mantissa, 274. 

 J. E. S." 



It is ohYiouslj A. purpureus, and is the specimen alluded to 

 by Bunge and Trimen, and correctly assumed by N. E. Brown to 

 be Linnteus's type of A. hypoglottis. It consists of two pieces 

 which may or may not be portions of one plant. Neither shows 

 anything of th^ root, and it could not be gathered from them 

 whether the plant is annual or perennial. The left-hand piece 

 has a head of about five flowers, the one on the right a head 

 of eight or nine well-developed legumes. 



