100 



KNOWLEDGE. 



[May, 1903. 



is occupied by a disk in the ceutre of which rise two styles. 

 The houey lies open to all comers. There is no guarding 

 of the approaches to tlie flower, or to the nectary ; no 



Fig. 5. — Secondary Umbel of Cow-Parsnep, showing large exterior 

 petals. Two-thirds natural size. 



provision for the visits of any particular kind of insect. 

 In consequence, all kinds of insects visit the flowers — flies, 

 beetles, bees, moths, ants — both flying and crawling 

 creatures. 



The umbels are generally terminal on the brandies, 

 but in a few cases axillary, as in the Water-Parsueps 

 {Helosciadium) and the Knotted Hedge-Parsley (Caucalis 

 nodosa). Compound umbels are found in the majority of 

 the British species, but sometimes the umbels are simple, 

 as in the Venus's Comb (Scandix Pecten-Yeneris). The 

 number of rays which go to compose the umbel varies 

 greatly, being reduced in some species to two or three, 

 thus forming irregular small clusters of flowers, instead of 

 large disk-like masses. The Marsh Pennywort, already 

 referred to, which has creeping stems from which arise 

 circular peltate leaves, has flower-stems much shorter than 

 the leaves, and quite hidden by them. These peduncles 

 terminate in little clusters of pinkish minute flowers. In 

 Astrant'iu, a quite abnormal effect is produced by the 

 petal-like character of the leaves of the involucre. The 

 involucre is the ring of small leaves which, more or less 

 developed, surrounds the point where the branches of the 

 umbel radiate from the stem. In some plants, as in the 

 Wild Carrot, this involucre is a conspicuous feature, 

 forming a ring of much divided hairy foliage ; in a few 

 species it is altogether wanting. In Astrantia the umbel 



Fig. 6. — Flower-head of Sea-Holly, with Involucre (lialf natural 

 size), and Single Fruit with Bract ( -x 2). 



is simple, with very numerous rays. The segments of the 

 involucre are longer than the rays of the umbel, and are un- 

 divided and erect, surrounding the inflorescence and giving 

 it the appearance of the flower-head of one of the Daisy 

 group of the Composit/e. In some of the sjiecies, as the pretty 

 A. carniolica, the colouring of the involucre adds to the 

 illusion. The Sea- Holly has also an abnormal inflorescence, 

 in which the primary umbel consists of a terminal shoot 

 and several spreading branches arising from one point, 

 and surrounded by an involucre of three leaves, while the 

 secondary umbels which these bear are reduced to egg- 

 shaped heads of flowers, with conspicuous spiny blue 



involucres, and also a spiny bract lielow each flower. The 

 Sea-Holly is quite one of the most remarkable as well as 

 one of the most beautiful of our British TJmhelliferee. Its 

 leathery, glaucous, spiny foliage, tinged with vivid blue, 

 and heads of blue flowers, are unique features among the 

 numerous native plants of this order. 



Of the characteristic fruit of this order something has 

 already been said. It consists of two single-seeded carpels 

 attached by their faces to a common short forked axis ; 

 when ripe the faces separate from the axis, the carpels 

 remaining attached and pendulous by their upper extremity. 

 The carpels are sometimes laterally compressed as in 

 Marsh Pennywort {see Fig. 1) ; sometimes dorsally com- 

 pressed as in the Wild Carrot (see Fig. 1) ; sometimes 

 smooth, often strongly ribbed ; sometimes covered with 

 spines. The fruit is usually devoid of any obvious 

 provision to assist its dispersal. When ripe and liangiug 

 lightly from the carpophore it is easily detached ; the 

 upright stems of the plants are often stiff and elastic, and 

 spring back into position if moved by a passing animal or 

 by the wind. The fruits may often by this means be 

 projected to some distance, in the manner of other cata- 

 pult-fruits. 



The fruits of some, as the Wild Angelica, are broadly 

 winged and may be carried to some distance by a high 

 wind ; but in very few cases is there any character which 

 suggests wind-carriage. Of more obvious use are the 

 hooked bristles with which the fruit of some of our 

 Umbelliferse is set. Some of the species of Bur-Parsley, 

 Caucalis, for instance, have their fruit thickly beset with 

 hooked bristles, liable to become entangled in the coats of 

 jjassing animals, and thus secure a wide dispersal. The 

 fruit of the Wood Sanicle is similarly furnished. A 

 curious case is furnished liy the Knotted Hedge-Parsley 

 (Caucalis nodosa), which bears small almost sessile lateral 

 umbels. The inner fruits of each cluster are smooth, 

 while the outer mericarp of the outer fruits alone is 

 furnished with numerous rows of hooked bristles which 

 are clothed with backward-pointing teeth. The fruit of 

 other species of Caucalis, and of the Wild Carrot (Laucus 

 Carota) possess an abundance of straight spines which may 

 act in the same ^vay as has been just described. The fniit 

 of the Venus's Comb (Sca7idix Fecten-Veneris) is very 

 curious. The mericarps and carpophore to which thev 

 are attached are prolonged into a great beak, which far 

 exceeds in length the fruit itself, attaining sometimes a 

 length of three inches. This beak is furnished with 

 forward-pointing bristles. The low growth of the plant 

 renders it unlikely that its dispersal is assisted by fruits 

 entangled in the fleece of animals, and the object of the 

 remarkable beak is not apparent. In the Sea-Holly, the 

 calyx limb is present — a very rare feature in this order — 

 as five stiff' lanceolate segments. These are persistent and 

 crown the rovigh fruit, and possibly iii some measure assist 

 its dispersal (see Fig. 6). But though the majority of 

 the Umhelliferm apparently trust to 'chance"' for their 

 advance into new ground, they still apparently secure a 

 sufliciently wide dispersal, as shown by the abundance of 

 many of the annual and biennial species. 



THE PAL^ONTOLOGICAL CASE FOR 

 EVOLUTION. 



By R. Lydekker. 

 (Continued from page 76.) 

 It will be well, before proceeding further, to sum up very 

 briefly the palseontological case for evolution so far as the 

 five main branches (or phyla, as Professor Osborn would 

 prefer to caU them) of the vertebrate stock are concerned, 



