i64 



NA TURE 



[Jan. I, 1874 



accessions of moisture that take place, by which this high 

 ice-cold region of the atmosphere over a great extent 

 and thickness, is brought to the point of saturation and 

 condensation. Underneath this leaden-hued mass of 

 cloud which uniformly covers the sky, but separated 

 from it by a clear space, is extended the dense cloud 

 covering of the paUio-cuinulits, which is formed by the 

 watery vapour of the atmosphere reduced to the points of 

 condensation and precipitation. This is the true rain- 

 cloiid, and it is fed and increased by the rapid drifting in 

 from below of torn masses of cumulus constituting the 

 fracio-cnmiilus or iL'ind- cloud . The fracto-cumulus may be 

 of all sizes, has no determinate shape, is the lowest and 

 swiftest moving of the clouds, and is whitish, greyish, or 

 slate-coloured, as may be determined by the hygrometric 

 condition of the air. On the return of fine weather acces- 

 sions of vapour by the fracto-cumulus slacken and then 

 cease, the pallio-cumulus diminishes in thickness and gra- 

 dually clears away, showing through its intervals the 

 pallio-cirrus above it, which in its turn is broken up, re- 

 vealing still higher up the delicate tracery of the cirrus. 

 The pallio-cirrus is negatively electrical, whilst the pallio- 

 cumulus is positively electrical, the clear stratum between 

 being neutral ; and between these oppositely electrified 

 strata, discharges frequently take place in thunderstorms. 



The merits of Prof. Poey's work are very considerable, 

 whether they be regarded as expository of Howard, or 

 as a contribution to this difficult branch of meteorology ; 

 and it is just those meteorologists who have paid particu- 

 lar attention to the observation of the clouds who will be 

 readiest to recognise its merits. It must, however, be con- 

 ceded that, as a descriptive classification of clouds, as well 

 as explanatory of the phenomena they present, Prof. Poey's 

 work leaves the subject in a state still too incomplete 

 to warrant us in recommending his system for general in- 

 troduction. It is a step in the right direction, and will 

 materially contribute to place this vitally important de- 

 partment of atmospheric physics on a satisfactory footing. 



Toward this end, what is now urgently wanted is an 

 extensive collection of the data of cloud-phenomena in all 

 countries, particularly of those clouds interesting in them- 

 selves or from their known relations to weather changes. 

 We have more than enough of unmistakeably pure typical 

 forms scattered through the pages of weather-literature, 

 but such do not greatly assist us. in describing and clas- 

 sifying many of the forms of clouds which occur. Hence 

 what is required is faithfully accurate delineations of these 

 forms in their different aspects, and systematic inquiries 

 set on foot into the relations of the forms of clouds to the 

 mode of their formation, to the states of the aqueous 

 vapour which compose them, and to the varying elas- 

 ticity, temperature, and electricity of the atmosphere. 



In connection with this part of the subject. Prof. Poey 

 investigated in 1862-64, by means of the ttiermo-electric 

 pile, the temperature of different parts of the sky under 

 different conditions, and of the clouds which passed across 

 it. Among other highly interesting results, he has shown 

 that the cumulus, properly so called, and the cumulo- 

 slratus of summer are the clouds of highest temperature ; 

 then follows the fracto-cumulus, except when it comes 

 after the rain which accompanies a thunderstorm, in 

 which case it is of a whitish colour, very rapid in its mo- 

 tion, much torn at the edges, and partakes of the low 

 temperature prevaihng on such occasions. The cirro- 

 cumulus is colder than the cumulus and the cirrus the 

 coldest of all the clouds. These are very suggestive re- 

 sults. We are convinced that the key to the position in 

 meteorology is a better knowledge of the vapour of the 

 atmosphere in its various states and changes ; and the 

 science will not make the advances it is destined to make 

 till meteorologists generally recognise the necessity of 

 equipping their first-class observatories with the requisite 

 appliances for carrying on those physical researches which 

 are intimately allied to meteorology. 



FERTILISA TION OF FLO WERS B Y INSECTS 

 V. 



More cotispicitons JJoiuers adapted to cross-fertilisation^ and less 

 conspicuous ones adapted to seif-fertilisation, occurring in di/ft- 

 rent species of the same genus. 



WHAT has been described in the two last articles as 

 occurring in varieties of the same species (using 

 the term " species " in its widest sense) we propose now 

 to investigate as existing likewise in species of the same 

 genus. 



Malva sylvestris and rotundifolia 



are two closely allied, but, as acknowledged by all 

 botanists, undoubtedly good and distinct species, differing 

 in their flowers in a manner similar to the two varieties 

 oi Lysimac/iia vulgaris and. the other species previously 

 considered. In both these species of Malva an oval mass 

 of anthers in the first place occupies the middle of the 

 flower, enclosing the stigmatic branches as yet undeve- 

 loped and lying close together (Fig. 23). At a later 

 period the stigmatic branches, growing out of and over- 

 topping the mass of anthers, spread and bend outwards 

 and downwards so as to occupy nearly the same place as 

 was before occupied by the anthers (Figs. 24, 25). Insects, 

 therefore, seeking for the honey which is secreted and 

 contained in five cavities between the lowest parts of the 

 petals (//, Fig. 23) and covered by a fringe of hairs (/r), 

 carry away on their hairy bodies the large prickly pollen- 

 grains from younger flowers, leaving many of them on the 

 stigmatic papiilasof thebranchesof thestyleofolderflowers, 

 which they can scarcely avoid grazing in seeking for the 

 honey. Hence, in both species, whenever insects frequently 

 visit these flowers, cross-fertihsation in the mannerdescribed 

 is largely effected, whereas self-fertilisation can scarcely 

 take place, neither spontaneously nor by means of insects, 

 nearly all the pollen-grains havijig been removed before 

 the unfolding of the stigmatic branches. Since, however, 

 Malva sylvestris and rotundifolia grow for the most part 

 in the same locality, and flower during several months 

 at the same time, insects flying about and seeking for 

 honey are much more likely to find out and visit the highly 

 conspicuous flowers of M. sylvestris than the far less con- 

 spicuous ones of .)/. rotundifolia J the former, when fully 

 opened, presenting bright rose-coloured bells of from 40 

 to 50 mm. diameter, the latter, on the contrary, light rose- 

 coloured bells of only from 20 to 25 mm. 



Direct observation, indeed, fully confirms this supposi- 

 tion, the flowers of M. sylvestris being always found in 

 sunny weather visited by a variety of insects, whereas 

 those of M. rotundifolia., especially when growing inter- 

 mixed with M. sylvestris, are commonly overlooked by 

 them all. Thus,duringthe sixlastsummers, I have observed 

 on the flowers of M. sylvestris and collected more than 

 50 species of insects, many of them very frequently (2 

 Lepidoptera, 3 Diptera, 5 Coleoptera, 40 Apida', some 

 Ichneumonidie) ; while in the same space of time 1 found 

 on the flowers of M. rotundifolia but 5 species (4 Apid;e, 

 I Hemipter), and those only in single or a few cases. 



It is evident from these facts, that wherever our two 

 species of Malva grow together in the same locality, M. 

 rotundifolia- would be rapidly extinguished, unless it were 

 enabled to produce seed by self-fertilisation ; M. sylivstris, 

 on the other hand, is so commonly visited and cross-ferti- 

 lised by insects that self-feitilisation, if it were possible, 

 would never be effected, or only exceptionally. Accord- 

 ingly natural selection must have preserved and accumu- 

 lated those slight individual variations oi Af. rctundi/olia, 

 which afford facility for self-fertilisation, whereas in J/. 

 sylvestris the possibility of sclf-fertihsation being quite 

 useless, might be lost, and, indeed, has been, completely 

 or nearly lost. Thus in the flowers oi M. sylvestris, when 

 precluded from the visits of insects by covering them with 



