THE PROBLEM OF THE BRITISH MARSH ORCHIDS 141 



ringed spots. Dr. Keller, to whom I sent a water-colour drawin"" of 

 this pl.uit, was extremely interested in it, as it had never been found 

 on the Continent, and at once confirmed the identification of O. lati- 

 folia as one of the parents. It is quite evident that by latifolia he 

 did not mean prcBtermissa — first, because two plain-leaved parent^ 

 could not endow their offspring with ringed spots, and, secondl3% 

 because a drawing of prcetermlssa was also sent to him, but he did 

 not suggest that as one of the parents. The hypothesis that a 

 plain-leaved crossed with a spotted-leaved plant will give rise to 

 a ring-spotted hybrid may possibly eventually prove correct, but at 

 present it appears to rest purely on conjecture. I have been so far 

 unable to trace a single instance in which a known hybrid between 

 parents of these classes has presented ringed spots, and of all hybrids 

 between unspotted ^9r<^ferw«/ss« and spotted maculata whioh I havy 

 come across not one was ring-spotted. 



There are thus two hypotheses to be investigated, i. e.^ 



(1) That there are only two British marsh orchids, O. incarnafa 

 .and O. prcdtf^rmissa; all other forms are hybrids between one or 

 other of these and O. maculata. Mr. Druee and Mr. liolfe both 

 appear to favour this viSw. 



(2) That there are three, O. incarnata, O. prcetermissa, and ring- 

 spotted O. latifolia. There are therefore six possible hybrids, viz. 

 (1) incarnata X maculata, (2) prcBteTmissaXmaculatay (3) lati- 

 folia X maculata, (4) incarnataxlatifolia, (5) incarnata.xprcBter- 



missa, (6) latifolia xprteter miss a. If 0. ericetorum Linton bo 

 regarded as a species, the number i§ increased to nine. 



It would seem that the bewildering variety of intermediate forms 

 found growing wild is more likely to result from the combinations 

 of a number of different factors, than from the crossing of only two 

 species (when incarnata is absent) or at the most thi-ee. 



It is suggested that the following points should be investigated, 

 in addition to the two named above : — 



(3) Are there any localities in which ring-spotted latifolia grows, 

 from yNhioh. ijrcBtermissa or maculata, or both, are absent? 



(4) Do prcetermissa and maculata grow together in any placp 

 where the ring-spotted plant is absent ? 



(5) If so, are hybrids present without ringed spots, and are they 

 jiumerous ? 



(6) Is there any locality in which ring-spotted plants and un- 

 spotted prcetermissa grow together, but where there is no maculata 

 in the neighbourhood ? 



(7) If so, are the ring-spotted plants identical in every other way 

 with those without spots on the leaves P 



If any or all of these questions' can be definitely answered, it 

 would probably throw much needed light on a difiicult problem. 



The most satisfactory thing would be for some of our younsfer 

 botanists to grow unspotted frcdtermissa and maculata, fertilize 

 the flowers of the former with pollinia from the latter (or vice 

 versa), and raise plants from the resultant seeds. It could then 

 be definitely ascertained whether such hybrids ever have ring-spotted 

 leaves. The experiment would take a few years to carry out, but it 



