166 



other extremity. I hope shortly to publish a detailed account of this 

 experiment, and shall therefore not give more exact particulars here. 

 I had expected to obtain primroses and oxlips, but had not antici- 

 pated the occurrence of cowslips also. It is true, the recorded expe- 

 riments of Herbert and Henslow might have led me to expect the 

 result which appeared ; but I may now confess a lurking suspicion 

 that some unascertained cause of error had been at work in their ex- 

 periments. And since Hooker, Babington, and other botanists still 

 continued to describe the cowslip and primrose as two distinct spe- 

 cies, I may presume that they were also sceptical on the point. Now 

 I can see only a choice between two inferences ; namely, that the 

 cowslip and primrose are a single species only, or, that one species 

 can pass into the other in two descents — the oxlip being the interme- 

 diate step. The experiments of Herbert and Henslow show the cows- 

 lip passing into the primrose in one descent. 



Festuca pratensis (Huds.) and Festuca loliacea {Huds.) — For half 

 a century past, it has been customary with British botanists, to de- 

 scribe the Festuca pratensis and F. loliacea as two distinct species. 

 The difference between them has appeared so strong in the eyes of 

 some botanists, as to warrant them in placing F. loliacea under ano- 

 ther genus (Brachypodium). In Steudel's Nomenclator, which bears 

 the date of 1841, they are entered as distinct species ; as also in the 

 Catalogue published the same year for the Botanical Society of Edin- 

 burgh. I had, however, seen some evidences that one could change 

 into the other, before the Edinburgh Catalogue was published ; and 

 in the same year of 1841, I brought a wild root of F. loliacea into my 

 garden. Though planted in close unworked soil, it had become a 

 large tuft by 1 843, and in the summer of that year it produced nume- 

 rous flowering stems. Some of the stems retained almost exactly the 

 character ("spiked raceme") which distinguishes the wild F. loliacea; 

 while ot]fers of them had so far assumed the branched or panicled in- 

 florescence of F. pratensis, that a botanist would assuredly have as- 

 signed them to F. pratensis, unless informed that they had been taken 

 from a root of F. loliacea, or shown the intermediate forms, which 

 were also produced from the same root. A root of F. pratensis, re- 

 moved into the same garden, became in 1843 rather less like F. lolia- 

 cea, than it was in its wild state ; but in the dry summer of 1844, some 

 of its panicles were reduced nearly into racemes. I have also seen 

 these two reputed species pretty closely connected in a series of wild 

 specimens, collected by Mr. Tatham, in the neighbourhood of Settle. 

 In this case, F. loliacea appears to become F. pratensis simply by 



