July, 1903.] 



KNOWLEDGE 



161 



hydrogen and oxygen combine to form water, the volume 

 of the latter is less than that of the gases, and it is 

 assumed that the atoun shrink, expelling the ether (or 

 heat) from their interiors. Prom this instauoe Mr. 

 Ilovenden generalizes, and obtains the law that when two 

 substances combine to form another substance, the original 

 volumes of the constituents lieing together greater than 

 the volume of the product, ether is expelled, yjroduciug a 

 rise of temperature ; when the original volumes of the 

 constituents were together less than the volume of the 

 jiroduct, ether is absorbed, and the surroundings are cooled. 

 In some cases where the reaction occurs in air, and the 

 results are not those antici]jated, it is claimed that the air 

 molecules are squeezed or expanded, as the case may be. 

 It is, however, easy to disprove this theory, in a manner 

 that leaves no loophole for escape. Take a cubic centimetre 

 of gunpowder, place it in a strong steel cylinder of (say) 

 liHMj cubic centimetres capacity, and thoroughly exhaust 

 the cylinder. On exploding the gunpowder the constituents, 

 which originally occupied a volume of 1 c.c , now occupy a 

 volume of 1000 c.c, so that, on Mr. Hovenden's theory, 

 each atom has expanded to lOOO times its original volume. 

 Nevertheless, the resultant gases are found to be heated. 

 The heat (or ether) absorbed by the atoms cannot have 

 been taken from the cylinder, since the latter is certainly 

 no cooler than before. Where, then, has it come from ? 



Thus we see that, starting from phenomena which 

 anyone may observe, Mr. Hovenden has reached conclusions 

 which, far from possessing an advantage over the accepted 

 theories, are positively at variance with well-known 

 phenomena. In criticising Mr. Hovenden's views, it has 

 been found unnecessary to appeal to any other theory, 

 except to show that the phenomena observed are capable 

 of au independent explanation. As a result, it appears 

 that if we possessed no physical theories at all, and were 

 offered the above theory by Mr. Hovenden, we should be 

 logically forced to reject it as inconsistent with itself, and 

 with the greater part of oui- experimental knowledge. 



FAMILIAR BRITISH WILD FLOWERS AND 

 THEIR ALLIES. 



By R. Lloyd Praeger, b.a. 

 IV.— THE COMPOSIT.E. 

 The Compositie form the largest Natural Order of flowering 

 plants found upon the globe. About one-tenth of the 

 whole phanerogamic flora may be referred to this group. 

 They are of world-wide distribution, though their number 

 varies largely in different countries. Thus Humljoldt 

 estimates that they form one-half of the jdiauerogamic 

 flora of the tropical regions of America, and Presl states 

 the same for Sicily ; on the other hand, they formed only 

 one twenty-third of the collections of plants made by 

 Smith in West Africa near the Congo. In the British 

 Isles they form about one-tenth to one-twelfth of the 

 flora, the uncertainty as to exact proportion arising from 

 the difference in the treatment which botanists ai'cord to 

 the hundred and more British " species " of Hawkweeds. 



The character by which these i)lants are most readily 

 recognized, and from which they derive their name, is the 

 grouping of their flowers into dense, usually flatfish, heads, 

 surrounded by a ring of small leaves which fulfil the func- 

 tion of the sepals of au ordinary flower. In my last article, 

 we saw how in Uie UmhaUifi'nr the numerous small flowers 

 are gathered together, so that the whole forms a most con- 

 spicuous inflorescence ; but in these plants each flower is 

 stalked, and each group of flowers (umbellule) usually is 

 again stalked; whereas in the Compusitm the flowers are 

 unstalked (sessile), and are seated tightly on the flattened 



top of the flower-stem (receptacle). A much closer 



approach to the Compositie we may study in the Scabiou.s 



family (Dipsaceie), of which the Devil's-bit (Scibioga 



succisa) and Teasel (Dipsacus sylvedris) are familiar 



examples. Here, as in the Compositse, the flowers are 



sessile and set on a swollen receptacle, and surrounded 



by an involucre of small leaves. 



From this ueai'ly related Order, 



GomposiUe may be i-eadily dis- y' 



tiuguished by the stamens, of which 



the anthers join together to form a 



ring round the style ; and by the y/ 



erect, not pendulous, ovule. That ff 



the flower-form of the Compositse is 



ail eminently efficient form would 



appear from the vast number of its j, j j^ 



species, and from the immense ])ro- iiower' and disk- 

 fusion of individuals — in our own flower of the Daisy 

 country, of the Daisy and Dandelion, (enlarged), 

 for instance. And truly, these flower- 

 heads strike one as being in the highest degree efficient. 

 Examine a common Daisy. The yellow button-like disk 

 is composed of a myriad of small perfect flowers, with 

 yellow five-cleft tulsular corolla, and ring of fused 

 stamens surrounding the pistil. Of calyx we find hardly 

 a trace ; the close jiacking of the flowers leaves no room 

 for it, and renders it unnecessary as a protective structure. 

 In the Compositie the calyx is usually reduced to a 

 few hairs, which often play a valuable part in aiding 

 seed-dispersal, as we shall see, by growing as the fruit 

 ripens into a feathery plume or pappus, which acts 

 as a parachute. The marginal or raij flowers of 

 our Daisy have no stamens — are female : and their 

 corolla is white, and greatly expanded in an outward 

 direction— rthe only direction in which there is room for 

 expansion. These ray-flowers, in fact, are largely useful 

 in advertising the otherwise inconspicuous flower-head. 

 Similar devices we have already noted in the wild Guelder- 

 rose ; and we may compare with these such flower-heads 

 as those of the little Cornel {Cornus suecica) and the 

 Astrantias, in which the flowers are surrounded by a ring 

 of coloured ?ertye^, which serve the same purpose. Finally, 

 our Daisy head is surrounded by a close-fitting double ring 

 of small leaves, the involucre, which encloses the whole in 

 liud, and plays the part that a calyx usually plays in a 

 single flower. 



This vast Order may l>e divided into several well-marked 

 groups. The Cori/mbifrrie include the genera which, like 

 the Daisy, have a disk and a ray. and other rayless allied 

 forms ; the Cijnarocephalex, which include all the Thistle- 

 like forms, have the style swollen, and no ray ; the 

 Liguliffonc, which form the best-marked group, have all 

 the ilnwers liijiiJttle (or shaped like the ray-flowers of the 

 l>aisy), as the Dandelion ; while the Lahiatittonv include a 

 few foreign genera, mainly South American, in which 

 corolla of the central florets is two-lipped. 



The Composit.v yield us some of our handsomest garden 

 plants, such as Chrysanthemums, Dahlias and "Cinerarias," 

 the latter l>eing cultivated varieties of species of Seiwciu, 

 introduced from the Canary Islands ; also the Sunflowers 

 and their allies, Asters, and various kinds of "Everlasting." 

 Among vegetables, the best known are Jerusalem Arti- 

 chokes (the tubers of Hfliaiithiis tiiberusiis), And Xrtn-hokea 

 (the I'eceptacles and involucral scales of Ciptara Scohjmiis). 

 Many species are aromatic, astringent, or narcotic, but for 

 so vast a number of plants, com])aratiyely few are used 

 by man. 



To sketch even the comparatively few British repre- 

 sentatives of the Composit(r within the limits of a single 

 article were an impossible task. The list of Compositx in 



