IS EUCALYPTUS VARIABLE 1 335 



speaking of the English language and of its marvellous 

 flexibility, he declared that "Any part of speech may be 

 used as any other part of speech." 



My studies of this genus have shown me that variation 

 exists in every species with which I am acquainted. Some 

 species are undoubtedly "stronger" than others, but the 

 more we collect and the more we observe, the more we find 

 old barriers between old species break down. With some 

 species one is inclined to say, "What character is constant! 

 there is no safety unless one keeps the type in sight just 

 as the mariner does the light of the light-house." To pursue 

 the simile further I am sure that the only way to avoid 

 botanical shipwreck is to stick to the type. 



III. Some Studies in Variation. 

 We have now arrived at the point when it will be profit- 

 able to consider specific instances of difficulties of classifi- 

 cation through variation. I would invite attention to a 

 paper 1 I have recently written to illustrate this point. In 

 summing up, I show that we have the following names for 

 the Gum- topped Stringybarks of Tasmania (which extend 

 into Victoria and Southern New South Wales), that is to say 

 for practically the same tree : — 



1. E. Risdoni, Hook. f. var. elata, Bentham. 



2. E. radiata, Hook. f. (var. 4), non Sieber. 



3. E. obliqua, L'herit. 



4. E. regnans, F.v.M. 



5. E. amygdalina, Labill. 



6. E. dives, Schauer. 



7. E. haemastoma, Sm. 



8. E. virgata, Sieb. var. altior, Deane and Maiden. 



9. E. oreades, R. T. Baker. 



10. E. Sieberiana, F.v.M. 



11. E. delegatensis, R. T. Baker. 



1 The Gum-topped Stringybarks of Tasmania ; a study in variation. — 

 Read before the Roy. Soc, Tasmania, 1902. 



