Gladiolus Studies — II 231 



The colors in the third class are the result of both cell-sap and plastid 

 colors. They are termed combination colors. Judging by cases of Gladiolus 

 primulimis hybrids noted, this species seems capable of altering the colors 

 in such a way that the bright reds are subdued to salmon, apricot, ecru, 

 and cream yellow, no doubt traceable to a dilution of the cell-sap reds 

 by the yellow plastid colors from G. primulinus. The resulting colors are 

 combination colors. 



Reciprocal crosses are crosses in which both the male and the female 

 functions are ser\'ed by each plant ; in other words, crosses in which each 

 parent is used alternately as a seed bearer and as a pollen producer. 

 Naudin (1866), in describing crosses between Datura fer ox and D. laevis, 

 says that the two groups of offspring of this reciprocal cross were so identi- 

 cally like each other that the two sets might easily be regarded as one. 

 In other words, either species could be alternated as pollen or as seed parent 

 without an appreciable difference in result. Dan\'in (1888) \sTites: 



" Hybrids raised from reciprocal crosses rarely differ in 



external characters." Colonel Trevor Clark foiind no difference in 

 reciprocal crosses between Begonia Dregei and B. heracleijolia, or B. 

 cinnabarina and B. Pearcei. 



With gladiolus the results seem to differ from the above-mentioned 

 cases, perhaps due to the extreme hybridity. Lemoine obtained Gladiolus 

 nanceianus by crossing G. Sounder sit and G. Lemoinei, G. Sounder sii 

 being the seed parent. The reverse cross gives many fine flowers, but 

 none so rich in color nor so characteristic in shape. G. Colvillei is the 

 result of crossing G. cardinalis on G. tristis concolor, and the reciprocal 

 cross is not mentioned as being identical. In R. T. Jackson's hybrids 

 between G. gandavensis and G. purpureo-auratus, the latter was used as 

 the male parent; the reverse order gave little success, but no notes were 

 kept. It is generally considered that G. gandavensis is a good seed parent, 

 as results were better when it was so used with G. purpureo-auratus, G. 

 dracocephaliis (figs. 28 and 29), and G. Saundersii. Perhaps in all these 

 cases the species were not pure types, but hybrids; in which event seed- 

 lings of the generation first obsen^ed would vary among themselves as 

 much as they would in reciprocal crosses. When one parent is stronger 

 or more vigorous than the other, obviously the stronger one should be used 

 as the female because of a supposed superiority for seed production. In 

 many cases a morphological characteristic causes an incompatibility 

 between the parents. The style of the pistil may be so long that the 

 pollen tube of another species cannot fertilize the o^alles. 



In A. E. Kunderd's mind the ideal was a strain of gladioli which should 

 have ruffled segments. For more than twenty-five years varieties sho\Nang 

 a tendency toward ruffling have been in existence, such as White Lady 



