June 1 6, 1881] 



NA TURE 



143 



(Linnean Soc. Journ. vol. xiv. p. 278) he had given a 

 series of experiments made on ants with light of different 

 colours, in order if possible to determine whether ants 

 had the power of distinguishing colours. For this purpose 

 he utilised the dread which ants, when in their nest, have 

 of light. Not unnaturally, if a nest is uncovered, they 

 think they are being attacked, and hasten to carry their 

 young away to a darker, and, as they suppose, a safer place. 

 He satisfied himself, by hundreds of experiments, that if he 

 exposed to light the greater part of a nest, but left any part 

 of it covered over, the young would certainly be conveyed 

 to the dark portion. In this manner he satisfied himself 

 that the different rays of the spectrum act on them in a 

 different manner from that in which they affect us ; for 

 instance, that ants are specially sensitive to the violet 

 rays. But he was anxious to go beyond this, and to 

 attempt to determine how far their limits of vision agree 

 with ours. We all know that if a ray of white light is 

 passed through a prism, it is broken up into a beautiful 

 band of colours — the spectrum. To our eyes it is bounded 

 by red at the one end and violet at the other, the edge 

 being sharply marked at the red end, but less abruptly at 

 the violet. But a ray of light contains besides the rays 

 visible to our eyes others which are called, though not 

 with absolute correctness, heat rays and chemical rays. 

 These, so far from being bounded by the limits of our 

 vision, extend far beyond it, the heat rays at the red, the 

 chemical rays at the violet end. He wished under these 

 circumstances to determine if possible whether the limit of 

 vision in the case of ants was the same as with us. This 

 interesting problem he endeavoured to solve as follows : — 

 If an ant's nest be disturbed the ants soon carry their 

 grubs and chrysalises underground again to a place of 

 safety. Sir John, availing himself of this habit, placed 

 some ants with larvas and pupje between two plates of 

 glass about one-eighth of an inch apart, a distance which 

 leaves just room enough for the ants to move about freely. 

 He found that if he covered over part of the glass with 

 any opaque substance the young were alwajs carried into 

 the part thus darkened. He then tried placing over the 

 nest different coloured glasses, and found that if he placed 

 side by side a pale yellow glass and one of deep violet 

 the young were always carried under the former, showing 

 that though the light yellow was much more transparent 

 to our eyes, it was, on the contrary, much less so to the 

 ants. So far he had gone in experiments already recorded ; 

 but he now wished, as already mentioned, to go further, 

 and test the effect upon them of the ultra-violet rays, 

 which to us are invisible. For this purpose, among other 

 experiments, he used sulphate of quinine and bisulphide of 

 carbon, both of which transmit all the visible rays, and 

 are therefore perfectly colourless and transparent to us, 

 but which completely stop the ultra-violet raj s. Over a 

 part of one of his nests he placed flat-sided bottles con- 

 taining the above-mentioned fluids, and over another part 

 a piece of dark violet glass ; in every case the larva: were 

 carried under the transparent liquids, and not under the 

 violet glass. Again, he threw a spectrum into a similar 

 nest, and found that if the ants had to choose between 

 placing their young in the ultra-violet rays or in the red 

 they preferred the latter. He infers therefore that the 

 ants perceive the ultra-violet rays, which to our eyes are 

 quite invisible. 



Now as every ray of homogeneous light which we can 

 perceive at all appears to us as a distinct colour, it seems 

 probable that these ultra-violet rays must make them- 

 selves apparent to the ants as a distinct and separate 

 colour (of which w^e can form no idea), but as unlike the 

 rest as red is from yellow or green from violet. The 

 question also arises whether white light to these insects 

 would differ from our white light in containing this addi- 

 tional colour. At any rate, as few of the colours in 

 nature are pure colours, but almost all arise from the 

 combination of rays of different wave-lengths, and as in 



such cases the visible resultant would be composed not 

 only of the rays which we see, but of these and the ultra- 

 violet, it would appear that the colours of objects and the 

 general aspect of nature must present to them a very 

 different appearance from what it does to us. 



Similar experiments which Sir John also made with 

 some of the lower Crustacea point to the same conclu- 

 sion, but the account of these he reserved for a future 

 occasion. He then proceeded to describe some experi- 

 ments made on the sense of direction possessed by ants, 

 but it would not be easy to make these intelligible without 

 figures. After detailing some further experiments on the 

 power of recognising friends, he gave some facts which 

 appear to show that ants by selection of food can pro- 

 duce either a queen or a worker at will from a given egg. 

 Lastly he stated that he had still some ants which he 

 had commenced to observe in 1874, and which are still 

 living and in perfect health ; they now therefore must be 

 more than seven years old, being therefore by far the 

 oldest insects on record. 



THE WEATHER AND HEALTH OF LONDON^ 



TO the statistician London affords materials for the 

 prosecution of many inquiries such as could not be 

 obtained from the statistics of any other city either in 

 ancient or modern times. Among the more important of 

 these inquiries are those which relate to questions sug- 

 gested by the enormous aggregation of human beings 

 over a limited area ^^hich London presents on a scale 

 absolutely unparalleled in the world's history. It is one 

 of these questions we bring before you this evening, viz., 

 the influence of the climate on the health of the people ot 

 London. 



The relation of weather to health is a question which 

 has engaged the attention of Dr. Arthur Mitchell and 

 myself for many years. In an early stage of the inquiry 

 our attention was mainly directed to Scotland, and more 

 particularly to the data supplied by its eight large towns ; 

 but it was soon found that, owing to the sparseness and 

 other conditions of the population, and to the fact that 

 the division of time into months only, adopted by the 

 Registrar-General for Scotland, they were not sufficiently 

 minute to show the true relations of weather to the 

 fluctuations of the death-rate through the year. In truth 

 it was only after not a little unsuccessful labour, and what 

 could at best be characterised as no more than partially 

 successful work, that we resolved eight years ago to open 

 the discussion of the whole subject by an exhaustive 

 examination of the meteorological and vital statistics of 

 London and London alone. More specifically our reasons 

 for the selection of London were that it aflbrded data from 

 (i) an enormous population spread over an area so limited 

 that it might be regarded as having one uniform climate 

 during each of the seasons of the year ; (2) full weekly 

 reports of weather and the deaths from the different 

 diseases ; and (3) returns extending over a sufficiently 

 long period. 



In the case of diseases such as diarrhcea and bronchitis, 

 which seem to be directly and immediately under the 

 influence of temperature, and such epidemics as scarlet 

 fever and whooping cough, the rate of mortality from 

 which is largely determined by season and weather, a 

 comparatively small number of \ears is required to give 

 a satisfactory approximation to their true weekly curve of 

 mortality. But as regards the great majority of diseases, 

 it quickly became apparent that a thirty years' average 

 was required in the construction of curves which could be 

 accepted as true "constants" for the diseases to which 

 they refer. The thirty years beginning with 1845 were 

 therefore adopted. An examination of the curves shows 

 that some of their striking features, particularly those 



■ Substance of a Lecture delivered at the Royal Institution, March 25. 



