382 ANALYSIS OF THE WORK OF CARL THUNJ5ERG. 



they could only be satisfactorily unravelled by examining his 

 specimens. For instance, Protca conifcm Thunb. is represented 

 by several mounted sheets, among which are included such dis- 

 tinct species as Lcucadcndroii uligiiioswii R. Br., L. strictum 

 R. Br., and L. adsccndcns R. Br. This is also a case, among 

 others, where Meisner includes one of Thunberg's names in his 

 list of insufficiently known species or species of which the genus 

 is doubtful. Similar examples may be found in the cases of 

 Protca couiosa Thunb. {Lciicadeudron ccmulum R. Br., L. platy- 

 spcrmum R. \^\^.), P. pallcns Thunb. {Lcucadctidron adsccndcns, 

 R. Br., L. dccurrens R. Br.), P. strobilina Thunb. {Lciicadoi- 

 dron spdlhiilafiun R. Br., L. ovale R. Br.), P. spcciosa Thunb. 

 {Pr&.ca spcciosa L.. P. Icpldoc^^rpodcndrvn L.), P. conocarpa 

 Thunb. ( Lcncospcninnn conocarpuni R. Br., L. attcnuatum R. Br., 

 L. cIHpticiini R. Br., L. (jrandiflornni R. Br.). Several other such 

 cases could be cited, but these will suffice to show how hopeless it 

 would have been to straighten out the synonymy without reference 

 to Thunberg's own specimens. Then, again, Thunberg several 

 times described the same .species under different names, e.g., 

 Leucospcrmiiin JiypophyUnni R. Br., was written up on some of 

 his sheets as Protca hyopJiylla, on other as Protea hctcrophylla; 

 Lcucadcndroii fnsciflora R. Br. was named Pratea torta and 

 Protca tcnulfolia; Lcucadendron Icvisanuni Berg, was written 

 up in his herbarium as Protca Icvisana, other specimens 

 as Protca liirsuta. There are other cases where the same con- 

 fusion occurs. It is not at all surprising that Aleisner was unable 

 definitely to place all Thunberg's names when Thunberg did not 

 distinguish the dift'erence in his own specimens. 



This brief account will show that Thunberg's work on the 

 South African Proteacecc was poor. Not only did he fail to dis- 

 tinguish specific difference in many of his specimens, but he 

 ignored the work of Niven and Brown published some years 

 before his " Flora Capensis," which should have enabled him to 

 distinguish at the least the genera represented in his herbarium. 

 Since this paper was read my attention has been drawn to 

 a publication by Juel* on Thunberg's herbarium, in .which the 

 author refers the published names of Thunberg's plants to the 

 present-day genus and species. As a list of the Proteacecc col- 

 lected and described by Thunberg is given by Juel, I have, with 

 the permission of the editor, withdrawn the list which was pre- 

 pared for the above short sketch, 



* H. O Juel " Plants Thunbergian.T. Arbetcn utgifna med under- 

 stod af Villiclm Ekmans Universitetsfond— Uppsala," pp. 262-270. 

 Uppsala, 1918. 



