INTRODUCTORY ESSAY. XV 



these are mundane, that is, found in all or most temperate or tropical climates, as Ranunculus, Cle- 

 matis, Senecio, and many Grasses and Ferns ; and we cannot yet tell whether the difficulties are greater 

 with them than with the more local or endemic genera, as Copromna, Celmisia, Alseuosmia, and Dra- 

 cophyllum. Of the mundane genera again, some are chiefly composed of species which are L 

 is the case with the three first mentioned), while of others the species themselves are widely distri- 

 buted, as those of Potamogeton, Lemna, and many Perns. 



The fact of a plant having a wide range implies its being exposed to climatic differences that 

 often induce change, and the consequent propagation of forms or races that cannot be recognized as 

 members of one species, without full series of specimens from many localities. If we allow a suffi- 

 cient time, it is quite reasonable to suppose that geological or other natural causes (producing a 

 change of climate) may isolate by sea or desert, or by the intrusion of stronger plants that monopolize 

 the soil, the outlying abnormal states of a species that was once uniformly spread over an area. 

 To connect those dissevered members is often a work of great difficulty, for individuals of such races 

 frequently retain their character even when they have been under cultivation for many years. 



Hybridization has been supposed by many to be an important element in confusing and masking 

 species*. Nature, however, seems effectually to have guarded against its extensive operation and its 

 effects in a natural state, and as a general rule the genera most easily hybridized in gardens, are not 

 those in which the species present the greatest difficulties. With regard to the facility with which 

 hybrids are produced, the prevalent ideas on the subject are extremely erroneous. Gartner, the 

 most recent and careful experimenter, who appears to have pursued his inquiries in a truly philoso- 

 phical spirit, says that 10,000 experiments upon 700 species produced only 250 true hybrids-}-. It 

 would have been most interesting had he added how many of these produced seeds, and how many of 

 the latter were fertile, and for how many generations they were propagated. The most satisfactory 

 proof we can adduce, of hybridization being powerless as an agent in producing species (however 

 much it may combine them), are the facts that no hybrid has ever afforded a character foreign to that 

 of its parents, and that hybrids are generally constitutionally weak, and almost invariably barren. 

 Unisexual J trees must offer many facilities for the natural production of hybrids, which, nevertheless, 

 have never been proved to occur, nor are such trees more variable than hermaphrodite ones. 



nothing in common but tlieiv facility for transportation. From the above-mentioned hypothesis it would hence result 

 that whilst the differences of one degree (specific) are small and inconstant) those of a higher degree (generic) are 

 great and trenchant. To a certain extent, however, these facts are not incompatible, for we can imagine a flora 

 wholly composed of a few genera as well marked (generically) as Coprosma and Alseuosmia, whose species may yet 

 be as imdefinable ; or again, species may be well marked, yet variable in characters which would in no one's opinion 

 be of specific value. '' 



* Hybridization as an agent in confusing species is a very favourite argument with those who are fond of 

 founding species on inconstant characters ; when shown a specimen combining two such spurious species, they at 

 once pronounce it a hybrid — a very simple way of getting rid of a difficulty. In Ferns, the most variable of all 

 plants, hybrids were once generally admitted to exist, but the observations of Suminski have led to the discovery 

 of their sexual organs, whose arrangement and structure seem to preclude the possibility of such a pheuomenon. 



f See his observations ou muling, Hort. Soc. Journ. vol. v. and vi. 1S50-1851. 



% Unisexual plants are very interesting in many points of view, and in none more than in the varying de- 

 velopment of the sexes according to circumstances. Observations ou this subject are very much wanted : it has 

 been stated to depend on local circumstances whether the seeds of a bisexual plant shall come up male or female ; 

 and the fact of both kinds of flowers, or even of hermaphrodite flowers, often occurring on a plant that usually per- 

 fects one sex only (as in the monoecious Hop-plant described by Mr. Masters in Gard. Chron. 1S471, shows that 

 we may even speculate on the possibility of dicecious plants having sprung originally from a siugle parent, whose off- 



