THE ORCHID REVIEW. 237 



so intimately blended that it is not possible to resolve them into separate 

 elements. This has been termed " Blended Inheritance." A second class 

 consists of hybrids in which the characters are divisible into regions, in 

 some of which are grouped the characters of one parent, almost unaltered, 

 while the remainder bear a similar relation to the other parent. This has 

 been termed " Particulate (or Mosaic) Inheritance." A third group 

 consists of those which closely or altogether resemble one parent, while 

 the other appears to be entirely or nearly unrepresented. This has been 

 termed "Alternative Inheritance." But these groups are not as strictly 

 definable as the above would indicate, for they grade into each other, and 

 are connected by various intermediate stages. 



How far intermediate hybrids are liable to revert to their parental 

 characters is very vaguely known, for in some cases, where it is thought to 

 have been observed, there are strong grounds for suspecting that the 

 reversion is due to intercrossing with one of the original parents. And at 

 all events there are various intermediate hybrids which are known to 



these plants to conform to the Mendelian conception shows on bow slender 

 a basis of fact the hypothesis stands. 



There are no hybrid Orchids, as far as we know, in which careful experi- 

 ments have been made to test the question of reversion, and it is doubtful 

 how far the few records on the subject are reliable. But let us take a supposed 

 case. Phragmipedilum (Cypripedium) X Sedeni is a hybrid between the 

 markedly distinct species P. Schlimii and P. longifolium (the result being 

 identical whichever is used as the seed parent). It possesses the most un- 

 mistakable combination of the parental characters, but according to the 

 Mendelian " law " the ovules and pollen grains ought to retain their original 

 characters, and if the flowers were self-fertilised the results should be com- 

 parable with those obtained by Mendel. We are not aware that the experi- 

 ment has been tried, but we do know the result of again crossing the hybrid 

 with both its parents. P. X Sedeni re-crossed with the P. longifolium 

 parent (it is immaterial for our purpose which was the mother) yielded 

 P. X calurum, in which a considerable return towards the characters 

 of P. longifolium is seen, and re-crossed with P. Schlimii yielded P. X 

 cardinale, showing a similar approximation to the P. Schlimii parent. The 

 secondary hybrids are as fairly intermediate as the primary one, and all 

 possess intimately blended characters, and yet, according to the Mendelian 

 conception, no new factor was introduced, and the ultimate germ cells of 

 P. X Sedeni should be in the same pure condition as in the original parents. 



The fact is, many characters cannot be classed under the headings of 

 dominant and recessive, and the dominant characters are not always the 

 same. An example from Mendel will give a clearer idea of his argument. 



