MENDEL S PRINCIPLES APPLIED TO ORCHID HYBRIDS. 628 



form the details of the analyses of the three single characters, side by 

 side, and under the respective numbers of the individual hybrids. 



In this way readers may compare for themselves my analyses with 

 the accompanying photographs, which show well the extent of the colour 

 areas and the forms of the markings, if not the actual shades of colour, the 

 last-named really being less differential and consequently of less importance. 



From the table it will be seen that the totals of the three single 

 characters added together give 76 BS + 71 BI, which, considering the 

 small numbers used, is a fair approximation to the equality which Mendel's 

 theory presumes. 



That is to say, the determinants in the germ-cells of SI are found in 

 equal proportions of S and I, according to the law of chance. As far as 

 these few experiments go, it may, I think, be fairly said that they confirm 

 Mendel's theory of the purity of the determinants in the germ-cells of 

 hybrids : that is, so far as practical results are concerned. The biologist 

 would further require these BS and BI forms to be self -fertilised for 

 several generations to see if they remain pure BS and BI in those 

 characters, as they appear to be on the surface. I have arranged to 

 carry on the experiment in this way, and have already a number of 

 seedlings up, but I fear that some time must elapse before they arrive at the 

 flowering stage. Those readers who read my first paper on this subject 

 (vol. xxvi. pp. 688-695) will be somewhat surprised at the results given 

 above, because at that time I said that I did not see how Mendel's theory 

 could be maintained with Orchid hybrids in face of the facts then before 

 me. 



At that time I was under the impression that the whole colour of the 

 dorsal sepal of the flower was a single character in the Mendelian sense. 

 A glance at the table will show that on that basis Mendel's theory 

 could not be maintained, because, if the above three characters were 

 taken together as a single character, the result would work out as follows : 

 Out of the 49 hybrids there would be 11 BS + 10 BI + 28 BSI, which 

 result would of course not be in accordance with Mendel's theory of the 

 purity of the determinants in the germ-cells of hybrids. A short time 

 afterwards I discovered that the apparent single character, i.e. the colour 

 of the dorsal sepal of the flower, was without doubt a composite one, made 

 up of three single characters, as detailed above. I therefore decided to 

 postpone the publication of this paper until the hybrids flowered again, 

 with the present result. 



Just a word or two, before I conclude, on the practical aspect of 

 Mendel's principles as applied to hybridisation in general and to Orchid 

 hybridisation in particular. While, from the biological point of view, we 

 cannot yet say with certainty that Mendel's principles are proved beyond 

 doubt, yet so far as the matter has been carefully tested in many different 

 genera, and by several independent observers in several countries, there 

 can be no doubt that there is a substantial agreement with the principles 

 laid down by Mendel. In these circumstances I think that the practical 

 hybridist would be wise to adopt Mendel's principles as a working 

 hypothesis and base his calculations upon it. If Mendel be ultimately 

 proved to be right, the hybridist wiU be justified by his results. If, on 

 the contrary, Mendel's theory has ultimately to be modified (it cannot be 



