90 



Garden and Forest. 



[February 20, 1889. 



pink, red, crimson, maroon, etc. — thus showing the Ganda- 

 vensis blood. In many the color was of a peculiar magenta- 

 pink, quite similar to that often seen in Magnolias. This 

 color existed in none of the Gandavensis parents and was dis- 

 tinct from any that I have seen in Gandavensis varieties. 

 Twenty-six per cent, of the hybrids were marked on the lower 

 segments by the characteristic shape of the blotch in G. ptcr- 

 pureo-auratus; but, instead of being dull maroon-crimson, in 

 most cases it was rich crimson, the color found at the base of 

 the blotches in the species parent. This is an interesting and 

 striking fact. Only eighteen per cent, of the hybrids had the 

 characteristic Gandavensis stripe on the lower segments, and 

 tifty-three percent, had markings which in shape were a com- 

 bination in varying proportions of the linear stripe of the 

 Gandavensis and the blotch-like stripe of Purpureo-auratus 

 Howers. The yellow splash at the margins of the maroon 

 blotches in G. pii7'pxireo-aiiratus is largely inherited by the 

 hybrids, but is often wanting. It was noted in nearly half of 

 the seedlings. Not infrequently the form of the yellow splash 

 is retained, but the color is replaced by white. The arrange- 

 ment of the segments in most of the flowers was that figured 

 in i?alone like the species parent; but exceptions were fre- 

 quent in which the arrangement A, found so predominantly 

 in the Gandavensis varieties, also existed, but always in a 

 few flowers only of a spike. The form in eighty per cent, 

 of the flowers was sub-open or flaring, being close to the 

 form which would be made by modeling a pattern half way 

 between the two parents in shape. Some were widely 

 flaring, remotely resembling a full blown Magnolia in shape. 

 A few flowers (eight per cent.) were deep bell-shaped, 

 like the Purpui^eo-auratus parent. A few others (nine per 

 cent.) were close to the Gandavensis parent in shape. The 

 size of the foliage and habit of the plants was half-way 

 between the two parents in ninety per cent, of the hybrids. A 

 few (ten per cent.) tended toward the greater size and rigidity 

 of the Gandavensis parents. The corms bear small corms at 

 the tips of the stoloniferous underground side-stems, thus 

 following the character of the species parent. The percentage 

 in which this character occurred was not ascertained, but it 

 was generally present. 



Out of 175 hybrids flowered, only three closely resembled 

 the Gandavensis parents, two were close to G.purpureo-auratiis, 

 but they showed the foreign blood by being shot with red. 

 Mr. Endicott, in a recent paper,* noted that he had only had 

 two plants showing pure blood from many hundred seedlings 

 raised from G. pwpureo-auratus. This statement, with my 

 own observations, leads to conclusions of great interest and 

 importance. It proves the complete union of tlie parents in 

 the hybrid offspring, there being very rarely or never indi- 

 viduals which copy either parent to the complete exclusion of 

 the other. It proves, also, the extreme potency of foreign 

 pollen on G. purpitreo-aiira.tus. This would not be expected, 

 as the pollen of its own species would be supposed, on all 

 common principles of fertilization, to be prepotent. 



In the hybrids, the shade of color, the crimson central 

 blotches and the flaring form of the flowers may be consid- 

 ered the new characters developed by the mixed blood. They 

 are the characters due to the cross, of which traces only can 

 be found in either parent. The hybrids, in a most remarkable 

 degree, were, roughly speaking, half way between the two 

 parents in form, coloration, marking, habit, etc. 



Mr. Charles Putnam, of Salem, raises many Purpureo-aura- 

 tus hybrids. He uses for the pistillate parent the variety 

 known as Lemoinii, which is itself a hybrid between G. pur- 

 pureo-auratus and G. Gandavensis. His staminate parents are 

 Gandavensis varieties. The seedlings which he raises are 

 largely very close to or nearly identical with Gandavensis va- 

 rieties, which is further proof of the complete union of parents 

 in the offspring, as his seedlings are three-fourths of Gandaven- 

 sis blood and only one-fourth Purpureo-auratus, and therefore 

 would be expected to lean toward the Gandavensis parent. 



Mr. C. Sander has originated a hybrid between G. dracoceph- 

 alus and G. Gandavensis, which promises to produce a line 

 of hybrids of much importance. 



G. dracocephaliis is a tall, strong-growing species with hood- 

 shaped, yellowish green flowers, netted with very small linear 

 red dots. The only approach to a stripe in the basal segments 

 is the omission of the red, leaving a pure yellowish green line. 

 The species produces very large bulbs and an abundance of 

 bulblets. With high culture I have grown this species and its 

 hybrids to the height of six feet. A few of the hybrids were 

 given me, and in 1885 seed was saved from which dui-ing the 

 past summer ninety-six seedlings bloomed. 



*"The Species of Gladiolus," by W. E. Endicott; Garden and Forest, No. 31, 

 page 363. 



The flowers were naturally fertilized, and, judging from the 

 results, probably largely by Gandavensis pollen. The type 

 species was fertilized with Gandavensis pollen, and perfect 

 seeds matured, but as unfortunately they were planted with 

 seeds from the hybrids, no notes can be given on them sepa- 

 rately. None of the hybrids approximate to the Dracocepha- 

 lus type, but many come close to the Gandavensis, indicating, 

 as was probably the case, that the blood was three-fourths 

 gandavensis. The great value of the hybrid is the strength, 

 size and floriferousnessof the plants and the unique form and 

 marking of the flowers, due to a combination of the characters 

 of both parents.* One of the most important features of these 

 hybrids is that a large percentage bear flowers entirely free 

 from stripes in the lower segments. This is a character due to the 

 Dracocephalus blood, and will be highly appreciated by Gladi- 

 olus-growers,becauseof the possibility it offersof obtaining flow- 

 ers of whole colors. Thecharacteristiccarminestripesof Ganda- 

 vensis varieties make often, and it might be said commonly, 

 an inharmonious contrast with the ground color of the flower. 

 Several crimson, pink and red Gladioli of this cross of various 

 shades, and of great beauty, bloomed, and were distinguishqd 

 by the fact that the colors were pure and the stripe wholly ab- 

 sent. This is suggestive, as other species having no stripes or 

 blotches, and desirable markings would doubtless tend to give 

 to hybrid offspring their special features, as is so clearly seen 

 in hybrids of G. dracocephalus and G. purpureo-auratus when 

 crossed with Gandavensis pollen. 



One seedling of much interest was produced. The pistil- 

 late parent was G. purpureo-auratus, the staminate a Dracoce- 

 phalus and Gandavensis hybrid. The result was a flower 

 combining the characters of all three parents in a marked 

 degree. The form was a combination of that characteristic 

 of the Purpureo-auratus and Gandavensis parents, with sharp, 

 pointed petals; color, dull magenta-red, bearing the charac- 

 teristic Purpureo-auratus and Gandavensis blotch in the three 

 lower petals. On the margin of the blotches, in the middle of 

 the two lateral petals, was a blotch of distinct mottled yellow- 

 ish green. In the centre of these blotches was a line of clear 

 yellowish green. This color and the mottling were the only 

 Dracocephalus characters noted, but they were so distinct and 

 are so characteristic of the species that no doubt can be en- 

 tertained of their origin. Four other Gladioli from seed of G. 

 purpureo-auratus were noted, which, in color and peculiar 

 hood-like form, showed the blood of the triple parentage. 



Tiie complete combination of parents in plant-hybrids is 

 shown in a marked degree in the fine hybrids sent out 

 by V. Lemoine, of Nancy, under the name of Montbretia 

 crocosmicsflora. I have grown five of his hybrids, and they 

 are all directly intermediate between the parents, which are 

 Montbretia Fottsii and Crocosmia [Tritojiia) aurea. The com- 

 bination is seen in the color, size, form and habit of the in- 

 florescence. Some swing toward one parent more, some 

 toward the other, and there is a brilliancy of color in several 

 of the hybrids far exceeding that of either parent. 



The hybrids of Gladiolus purpureo-auratus and G. Ganda- 

 vensis and of G. dracocephalus and G. Gandavensis zxq entirely 

 fertile, and the seed grows easily. The Montbretia hybrids, 

 above noted, also seed freely, so that there seems here to be 

 none of that whole or partial sterility attributed to true species- 

 hybrids in plants and animals. 



The pollen of G. purpureo-auratus is much less active on 

 Gandavensis varieties than the reverse cross, and attempts to 

 make this cross were attended with little success. The same 

 is true of crosses tried with dracocephalus pollen. In a lot of 

 about 250 Gandavensis seedlings, bloomed for the first time 

 last summer, only three or four showed signs of Purpureo- 

 auratus blood, although the flowers had quite as good a 

 chance to be naturally fertilized by Purpureo-auratus pollen 

 as did tliat species with Gandavensis pollen. Insects must 

 have passed freely from one species to the other in order so 

 effectually to cross-fertilize the Purpureo-auratus flowers. No 

 Gandavensis seedlings showed traces of Dracocephalus bloodr 



To sum up the deductions. The species G. purpureo-aura- 

 tus is very readily fertilized by Gandavensis pollen. The off- 

 spring of G. ptcrpureo-auratus and G. dracocephalus crossed 

 with G. Gandavensis ave vigorous and readily set fertile seed. 

 The hybrids are a direct combination of the two parents in 

 color, form, habit and root, but with some new characters 

 developed by the mixed blood. 



It is sometimes thought, with garden-plants long in cultiva- 

 tion, from which seedlings have been raised in abundance 

 and for a considerable time, that we have reached the end of 

 the striking variations possible. Such a statement has been 



*A few of these hybrids were shown at the Massachusetts Horticultural Society 

 last August, and were noticed in Garden and Forest, No. 28, p. 336. 



