February 20, 1S89.] 



Garden and Forest. 



89 



laterals ; the lowermost overlaps its laterals ; the lower 

 pair of laterals is wholly within, and the upper pair oi laterals 

 is wholly without 

 their adjoining seg- 

 ments. A single spike 

 may be composed of 

 tiowers of the first ar- 

 rangement(^)wholly, 

 or it may have flowers 

 of both arrangements 

 in varying numerical 

 j)roportions ; but the 

 first (A) as far as noted 

 always predominates. 

 Flowers of the second 

 arrangement {!>) may 

 be the first, last, or 

 s c a 1 1 e r i n g 1 y inter- 

 mediate on the spike. 

 The two arrange- 

 ments are funda- 

 mental in the flower, 

 they are not brought 

 about by twists in the 

 segments. The ar- 

 rangementof the cells 

 in the ovary coincide's 

 with the varying rel- 

 ative position of the 

 segments. 



hi G. dracocephalus 

 all ihe flowers, in the 

 large number exam- 

 ined, were arranged 

 as in B. The same is 

 true of G. purpureo- 

 aitratiis. Only a few 

 flowers of G.Saicnder- 

 sii were observed 

 when this character 

 was in mind; but 

 those were of the 

 second arrangement. 

 G. psittacinus is the 

 only species where a 

 variation of the ar- 

 rangement was noted. 

 Int his species most 

 of the tiowers were 

 arranged as in A; 

 but a very few were 

 found of the type of B 

 in the diagram. 



The existence of 

 two plans of flower- 

 arrangement on a 

 single spike in a true 

 species would be ano- 

 malous, and I con- 

 sider its occurrence 

 in hybrid Gladioli due 

 to the inheritance of 

 mixed blood ; the ar- 

 rangement (A) being 

 due to certain ances- 

 tral species, and the 

 arrangement {B) to 

 other ancestral 

 species. This sugges- 

 tion is borne out, in 

 fact, by the hybrids of 

 G. purpureo-aiiraius, 

 for, as noted below, 

 both arrangements 

 are found in the 

 hybrids, although 

 the form B only is 

 characteristic. It is 

 quite possible that, by 

 following this charac- 



96. — Gladiukis purpur 



terout one might be able to trace the origin and kinship of 

 species and garden-hybrids of Gladioli. 



The species Gladiolus purpiirco-auratus, used as the pistil- 

 late parent, is well represented in the illustration upon this 

 page. Theflowersaredeep, tubular, and bell-shaped. The ar- 



rangement of the segments is as in 5. On the central lower petal 

 is a narrow blotch, and on the adjoining laterals broad, elongate 



blotches of dull ma- 

 roon-crimson. At the 

 base of the blotches, 

 near the centre of the 

 flower, the color 

 heightens to a rich 

 crimson over a very 

 small area. This ricii 

 color is an important 

 fact in the result oIj- 

 tained in the colors 

 of the hylirids, where 

 the whole blotch is 

 commonly a rich 

 crimson. On the mar- 

 gin of the blotches is 

 a lip-like splash of 

 golden yellow (not 

 shown in the figure). 

 The color of the rest 

 of the flower is a pale 

 greenish-yellow. The 

 plant grows to the 

 height of about four 

 feet, with foliage nar- 

 rower than in Ganda- 

 vensis varieties. The 

 flower-spike grows in 

 a graceful svibarcuate 

 fashion, quite distinct 

 from that character- 

 istic of Gandavensis 

 varieties or of any spe- 

 cies with which I am 

 acquainted. The 

 flowers are arranged 

 rather far apart on the 

 stem; they all face in 

 one direction, and in 

 view of the arcuate- 

 ness of the flower- 

 stalk, and pendant, 

 bell-shape of the 

 flowers, their interior 

 isnoteasilyseen. The 

 cormsof G.piirpureo- 

 auratus are not very 

 large. They bear bulb- 

 lets at the base, and, 

 in addition, small 

 corms at the tips of 

 the characteristic run- 

 ning root-stocks, one 

 of which is shown in 

 the figure. The spe- 

 cies is distinct, and 

 different individuals 

 vary little. The only 

 variation noted was s. 

 specimen which had 

 a richer yellow than 

 common in the lip 

 borderingthe maroon 

 blotches. 



The Gandavensis 

 varieties used as the 

 staminate parents 

 were a choice lot of 

 named varieties ani.l 

 seedlings. 



The quality of the 

 Purpureo-aiu'atus hy- 

 brid seedlings was 

 of a high order. Of 

 the 175 which bloom- 

 ed, ten per cent, were 

 plants worthy of a 

 place with the best 



.See pat; 



named varieties. Thirty-seven per cent, were well worthy of 

 general cultivation. Fifty-three per cent, were of medium 

 quality, indifferent or poor. Inestimatingquality, colorwas con- 

 sidered of first importance, and, after it, form and arrangement. 

 All the flowers, without exception, were colored — white. 



