107 
The Oncocyclus Section 
Capsule , small, not much more than £ in. long, the outline being a pointed oval, the section 
trigonal with three sharp angles ; it dehisces completely while still somewhat green and while still 
hidden in the persistent spathe-valves. 
Seeds, small, brown, smooth, oval or globular, with curious, transparent, almost gelatinous appendages, 
often longer than the circumference of the seeds and twined round them. These appendages quickly 
shrivel on exposure to the atmosphere. 
Observations. 
This very distinct Iris has been known since the middle of the eighteenth century. There exists 
at the British Museum a specimen from Bartram’s Herbarium, dated 1764 and described as "a sweet- 
scented plant, growing 5 in. high, which spreads much and differs from the Carolina dwarf Iris." 
The "Carolina dwarf Iris” is probably a reference to I. vema, with which I. cristata was at first 
confused, e.g. among Pallas’ specimens (BM) which are labelled either vema or even pumila. 
This confusion is hardly a matter for surprise in view of the fact that in Kentucky, at any rate, 
cristata and vema are sometimes found in company, though vema is usually found at a greater elevation 
than cristata (cf. note on Short's specimen (K)). 
The points of agreement between I. cristata and I. lacustris are so many and they are both 
separated by so many characters from all other Irises, that it seems impossible to give both specific 
rank. It is true that /. lacustris is usually smaller than /. cristata but this character tends to disappear 
when seedlings are raised, though even in seedlings the closer growth of the more slender rhizomes 
is still apparent. However, there is no difference except size in the growth of the two plants and 
size alone is often a matter of soil and climate. Moreover, as the smaller plant comes from the colder, 
northern region, we are surely justified in looking upon I. lacustris as a mere local form of I. cristata. 
As garden plants, both do well in moist soil composed of humus and gravel, in the kind of soil, 
in fact, in which they grow in their native home on the banks of streams. Both are quite hardy and 
I. lacustris has, or, at any rate, some plants of it have, a curious habit of flowering at odd times from 
May until October. 
Propagation is easy and is best carried out by cutting away the side-growths soon after the 
flowering season. The points of the new roots will then be apparent and each of these stolon-like 
growths is capable of becoming a flowering plant by the next year. Each rhizome that has flowered 
withers and dies and may therefore be at once discarded, for no fresh lateral growths will make their 
appearance from it. The plants benefit greatly by this annual remaking of the plantations, and the 
opportunity should be taken of adding fresh supplies of well-decayed leaf-soil. 
I. cristata does not produce seed very’ readily in cultivation in this country. Even when a few 
capsules are obtained as the result of artificial pollination, the seeds in each are not numerous. More- 
over, germination appears to be difficult, for though I have regularly sown seeds of /. cristata for 
several seasons past, none have ever germinated except a few from one pod of the variety lacustris. 
An albino form of this Iris has recently appeared in commerce. It is uncertain whether /. cristata 
alba was found wild or whether it is of garden origin, although there is reason to believe that it has 
been found among collected plants. 
THE ONCOCYCLUS SECTION 
This group is probably the most difficult of all for botanists and gardeners alike. The former can 
only in the case of very few species find specific differences that can be recognised in herbarium 
specimens and the latter finds it almost impossible to induce the plants to live on from year to year 
in our capricious climate. 
The meaning of the word Oncocyclus is not apparent, for Siemssen, its inventor', contented himself 
with the remark that it is composed of the words oyxos and kvk \ o s but vouchsafed no indication of the 
meaning that he attached to the compound. We may conjecture that it may have had some reference 
to the character of the seeds, see Plate XLV 1 II, Fig. 8, which are distinguished from those of all the 
other groups by the large creamy-white aril, which is sometimes as large as the seed itself. All the 
Oncocyclus species or varieties agree in the possession of seeds of this type and, except perhaps in size 
there is no apparent difference by which to distinguish those of any one of them. 
The other characters which all Oncocyclus Irises have in common are : 
I. The bright red-skinned rhizome with crowded growths with occasional and irregular extensions 
by means of stolons (see Plate XLVII B representing an extreme example of this system of growth 
in the Regelia 1. stolonifera). 
’ Cf. BZ. 1846, p. 705. 
14—2 
