215 



ILES STERILES. 



I am indebted to Dr. L. Diels for tlie following note : — 



As suggested by your remark on p. 1 I 7, Part IV, I avail myself of the opportunity to set forth my 

 views about the locality " lies >Steriles " recorded by Baudiu's Expedition. " Having gone through several 

 of the old original books, I am satisfied (with you) that this name has never been used in published 

 literature. At the same time, there is mi doubt to me that it is a translation of the Dutch "Dorre Eylandt" 

 (barren island), and means, in a broadei- sense, those three islands called nowadays Dirk Hartog Island, 

 Dorre Island, and especially Bernier Island, in Shark's Bay. These islands, being discovered by Dirk 

 Hartog in 1616, were more thoroughly explored, for the first time, by the e.xpedition of Baudin. They are 

 fully described in Perron's and lj. de Freycinet's report of this voyage (" Voyage ile decouvertes auxTenes 

 Australes," Paris). There is a quite detailed paragraph on their vegetation in this book. It is safe to 

 suppose that several species have been collected on Bernier Island. I think the species labelled "jles 

 Stei'iles " came from there ; for all of them wo are aware of belong to the flora of sandy dunes on limestone 

 formation, just as it is met with on these islands ; for instance, Eucalyptust fiwMndu, which was collected 

 again near Shark's Bay by Milne (v. p. 1 15 of your Revision) ; further Beyer'm ryitnrt<<rn.n, Bth. (F/or. Ausf 

 vi, 66), this plant has been collected again or Dirk Hartog's Island by Naumann (in Herb. Berlin) ; and 

 even more deciding, Srholt^w Irplantha, Bentn. "We have this plant from " lies Steriles " in Herb. Berlin, 

 communicated by the Paris Museum, as to herb. R. Bi'own (vide Bth. F. Au«(., iii, 70). Now the same 

 species was collected near Shark's Bay by Milne, on Dirk Hartog's Island by Naumann, on dunes near 

 Carnarvon by my.self. The whole evidence leads me to the conclusion that " lies St^riles " are those (really 

 exceedingly barren) islands in Shark's Bay. The name, then, is an extension of the old Dutch " Dorre 

 Eylandt," which meant only one of them. That this informal, rather provisional naming has been retained 

 on the labels, while the official report has only the valid names (lie Dirk Hartog, Ile Dorre, Ile de 

 Bernier), is not sur|)rising when one considers how very little care was taken about correct labelling bv 

 the old botanists. 



Explanation of Plates (69-72). 



PLATE 69. 



E. oli^OKK, F.V.M., var. Flocktoni, Maiden. 



\(i, Ml. Mature leaves; Ic, buds; \d, \r, fruits of the type. Desmond, near Ravensthm-pe, Western 



Australia. (.J.II.M.) 

 '1(1. Mature leaf; ih, bud and llowcis ; ir, anthers; J(/, fi-uits from co type. Ksjierancc, Western 



Australia. (L. L. C'owen.) 

 .■)r(. Matuic leaf ; .'}/), fruits much h^ss corrugated tliaii those of the type. iMurcliison Bivrr, ^^'esterM 



Australia. (Oldtield.) 

 l'^ Fruit, smaller in size and 1('.5S conugated ; l/i, anther. Coweowiug, M'estern Austialia. (Max Ivoch. ) 

 I'igures '^ and 1 are co7inectimg links between K. oleoaa and var. Flock/otii, and form part of the evidi>nc<- 



that the latter form, dissimilar as it looks at first .sight, cannot be given specific rank. 



A'. Soiirfii., II. sp. 



.")«. Juvenile leaf ; -jh, intermediate leaf ; -"ic, mature leaf. Kalgoorlie, Western Australia. (J.H.M.) 

 6(^ 6/). Buds ; 6r-, Viil, fruits. Nearly 70 miles north of Kurrawang, Western Au»trali (J.H.M.) 



7. AnthiT, from :i tree near K iirr.-iwang. (.I.H.IM.) 



E. C'Miinrli, n. sp. 



?<(/. -luvenile leaf ; 86, intermediate leaf; Sc, mature leaf; f<d, buds .-, *iUfts of the type. Goongarrie, 

 Western Australia. (J.H.M. 



