174 THE JOURNAL OF BOTANY 
applied to his “species no. 2.” We think a eh sea is 
poor evidence that Linnzus did confuse tet iti under his 
m 
sum corniculatum luteum, minus repens & etiam procumbens 
M. H. 2, 184,” i.e. replacing Morison’s name by its more modern 
a 
equivalents, O. corniculata auc O. repens Thunb. etiam 
Robinson. 
In an case, it is incorrect to state ai “According to the 
e 
supplementing the published matter. In this case the published 
matter is quite definite: Oxalis caule ramoso, pedunculis multiflors, 
is the “creeping prostrate-stemm ee Pek t 
only i is the Herb. Cliffort specimen referred t “type,” but it 
is merely a misidentified specimen. On the othe “hand, there are 
i bi us 
Cliffortianus, the corresponding specimens in Cliffort’s herbarium 
being sorneotly ae rded as types. 
ut in cases where the names in the Hortus Cliffortianus 
refer to a composite collection of “eteres nees to various authors, 
none of the specimens can be regarded as t es, and it is the 
references alone which matter. Mr. Lacaita, in this Journal 
(1912, p. 223-4), has already pointed out that Linneus probably 
never again saw Cliffort’s herbarium after he finished the Hortus 
Cliffortianus. Evenifthe International Rules were not so definite 
as they are, one would be compelled to regard the references from 
the Species Plantarum to the Hortus Cliffortianus as mere ely 
references to the published matter of the book. In those cases in 
which the pee refer to a list of references, the herbarium is 
sometimes composite as the book—there one a inher 
specimens represeliting the species under various of t mes 
cited, and sometimes even under names which do not sett in 
book. This is the case with this Ovzalis, since besides the 
specimen referred to above there is another labelled ‘‘ Oxys Lutea 
J. B. 2. 388.” This is a characteristic specimen of O. corniculata 
LL. et auct., showing how impossible it is to regard any such 
specimens as types. The conclusion arrived at would depend 
upon which specimen was seen, and might be changed: if one 
specimen were ever lost or missing or not collected from the 
garden to be placed in the herbar 
here is therefore no reason for | the changes in nomenclature 
proposed by Dr. Robinson, see: the principle of residues cannot 
be applied to the case of O. corniculata L., which is not composite. 
In conclusion, esire to ratte on the necessity of being very 
ore using any of these old specimens to decide ef sae 
in the fase of other evidence) what species must bear th 
given in the Species Plantarum. 
