460 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



We may well believe Green's statement that seeds taken from 

 singles appearing among a large number of doubles were found to give 

 in turn many doubles, though the explanation of this fact is not, as he 

 supposed, to be traced to the close proximity in the ground of the two 

 kinds of plants. We find the same instruction in Paxton's " Botanical 

 Dictionary " in 1868. He says : "In order to obtain good double 

 Stock-gilliflowers, Brompton, and Queen Stocks, choice should be 

 made of such single-flowering plants as grow near many double 

 ones." * Only a year or two ago I myself heard this view strongly 

 maintained by a well-known gardener. But although this sug- 

 gestion has been constantly repeated by various writers on the 

 subject, they have generally admitted that the occurrence of 

 double flowers with pollen must nevertheless be extremely rare. 

 In fact the impression made on the reader is that the writer is 

 often handing on a tradition, and in no way vouching for the 

 occurrence upon the supposition of which the practice is based. The 

 nearest approach to first-hand evidence that such flowers occur which 

 I have been able to find is contained in the work by Phillips (1824) 

 to which I have already referred, t Having stated that the full flowers 

 never produce seed, he adds : " Yet it is the opinion of most gardeners 

 that they assist in causing the plants to become double, and it will be 

 observed that there is frequently a straggHng anther to be found in the 

 double blossoms, which may assist this change by the impregnation of 

 the neighbouring plant." He does not, however, mention whether 

 this anther-hke structure had been observed to contain pollen, and if so, 

 whether the pollen appeared to be good. As to both points I think 

 we may feel considerable doubt, but neither need detain us here. For 

 even if such double flowers are occasionally produced, they must be 

 so extremely rare that their existence could have no appreciable effect 

 on the constant production of the large and definite proportion of 

 doubles which we observe. I may add that I myself have never seen 

 any structure remotely resembling an anther in the many hundreds 

 of double flowers which I have examined. A correspondent writing 

 to the " Gardeners' Chronicle " in 1843 says that he found eminent 

 Stock-growers were not aware that the doubles had no pollen, but 

 that when asked by him to examine a flower they agreed that such 

 was the case. J Here, as in the case of the previous method, the very 

 plausibility of the view seems to have led to its ready acceptance 

 without further verification. 



5. The selection for sowing of seeds which are small and irregular. 

 (See a paper by Thiele (1825) quoted by Goebel,§ Regel || (1852) , 

 " Gardeners' Chronicle " ^ (1859). ) 



* New edition. f See p. 29. f p, 678. 



§ Pringsheim's Jahrbuch, Bd. 17, 1886. 

 II Gartenfiora. 1852, p. 84. 



^ Pp. 51, 52, where an account is given of a paper, translated for the Horti- 

 cultural Society of Paris by M. Courtin, which had appeared in the Ilhistnrie 

 Garten-Zeitung. According to the author, the best seeds are those which have a 

 very^small border, but which are irregular in form, and rather angular, not flat. 



