shown fail to recognize the bird at all. They speak also of another flightless bird by the name of Mehunui. This bird, one old and very intelligent Moriori informed me, was the same as the Maoris called Kakapo. Mr. Alexander Shand, an old resident in the Chatham Islands, and the sole European living versed in Moriori customs and traditions, and capable of speaking their language with fluency, also confirmed this information, and told me that the Kakapo (according to the Morioris) was very abundant in the islands prior to 1836. He himself in the early days had seen their burrows often. I had observed, while collecting, several Psittacine bones, and on learning this fact I felt sure that those I had picked up and packed away must belong to Stringops. On my arrival here, however, I find so far that there are no Kakapo bones in the collection, the Psittacine bones being the head and beaks of Nestor notabilis, the Kea. I have as yet had time to do no more than run through the collection I have brought back; but there appear to be in it several large Ralline tibiæ of species unknown to me. I am looking forward to another opportunity of thoroughly exploring these interesting islands with more time at my disposal than I could afford on this occasion. HENRY O. FORBES.

Canterbury Museum, February 23.

Pigments of Lepidoptera.

A LETTER of mine on the subject of butterfly pigments was published so recently in NATURE (December 31, 1891, p. 197) that I hesitate to ask for further space at the present time. But the appearance of Mr. Perry Coste's articles, together with the tone of some remarks made by him at the close of the last article, lead me to venture upon a few words, partly in criticism of a theory he advances, and partly (though this is less important) in claim of priority, since Mr. Coste does not do me the honour to refer to my work on the subject.

The chief generalization which Mr. Coste bases upon his experiments is that which he terms the "reversion effect," that is to say, the production of yellows from reds by the action of acids, and the restoration of the former by neutralization and other means. The theory that he advances to explain these phenomena-namely, that the red body acts as a base, and forms with acids salts which are yellow-is quite untenable. As I have shown (Proceedings of the Chemical Society, June 1889; vide NATURE, vol. xl. p. 335), the soluble yellows are themselves acid bodies of quite definite composition. Indeed, as far back as 1871, Prof. Meldola called attention to the fact that the pigment of G. rhamni was soluble in water, and showed that its aqueous solution had an acid reaction. Mr. Coste has worked with D. eucharis; if he will dissolve the red pigment from the border of the hind wing of this insect in pure water, he will find that a yellow solution is the result, but that, if the solution be evaporated to dryness, the solid residue of pigment is red once more; showing that there is either the question of hydration to consider, or a weak combination of the yellow acid body with a base, which is dissociated in aqueous solution. At any rate, I have obtained from this red pigment of eucharis a silver compound which contains a percentage of metal exactly equal to that from the pigment of G. rhamni.

In 1889 I was able to predict possible constitutional formulæ for the acid yellow pigments, and am happy to say that recent careful combustions of their silver salts to a large extent confirm my original ideas. My results will be shortly published in

extenso.

Mr. Perry Coste's experiments are very useful as forming a method of classifying these lepidopterous pigments; but, if he will forgive me for saying so, they are of far too empirical a nature for any considerations as to the constitution of the bodies to be based upon them. As one who has for many years past spent a large portion of his time and no inconsiderable portion of his substance in obtaining a sufficiency of these pigments for analysis and investigation, Mr. Coste will forgive me if I do not respond to his invitation to leave him to continue his researches alone." It is hardly well for one investigator to say "hands off" to another, and I shall not imitate Mr. Coste in this matter; but will only express a hope that in his future work he will not confine himself to the immersion of the wings of his insects in strong and destructive reagents.

I have lately been working at the genesis of these pigments in the pupæ, and might say something with regard to the nature of

THE new edition of Prof. Everett's "C.G.S. System of Units" contains, at the very beginning, two misleading statements, based seemingly on a misapprehension of facts. In so valuable a work, such errors are to be deplored.

Pp. xiii. and xiv. give an account of weighings made at the International Bureau of Weights and Measures between certain "standard pounds" and the international standard of mass. From the statement as given, it would be inferred that there is room for doubt as to the relation between the British standard of mass and the international kilogramme.

The real facts are, that the standard pounds were only nominally "pounds"; they were standards with known corrections, which, however, have not been applied to the equivalents given on p. xiv.

The true relation of the Imperial pound to the international kilogramme is given in the Board of Trade Report of Proceedings under the Weights and Measures Act, 1884 (p. 4), according to which the Imperial pound = 453'5924277 grammes.

On p. 34 of "C.G. S. System of Units," Mr. Chaney's results of the weight of a cubic inch of water are discussed, and the conclusion is reached that Mr. Chaney's result differs by o'00125 from the theoretical relation between volume and mass of water in the metric system.

This result is obtained by comparing Mr. Chaney's results, without reduction to vacuo, with the mass of a cubic centimetre of water.

Mr. Chaney states that the standard air to which his result is reduced weighs 0'3077 grains per cubic inch. Therefore his result reduced to vacuo is: one cubic inch of water at 62° F. weighs 252 286 +0.308 = 252.594 grains.

If we take the value for the thermal expansion of water, in terms of the hydrogen thermometer scale, as determined at the International Bureau, we find the density of water at 16° 667 C., = 62° F., referred to its maximum density o'998861.

=

=

Using the equivalents I metre = 39'3700 inches, and I gramme = 15 43235639 grains, we obtain: one cubic inch of water at 62° F. weighs in vacuo 252 6045 grains; while Mr. Chaney found 252.594 as above given, a discrepancy of one part in 25,000 only, as compared with one part in 800, given by Prof. Everett. It is not clear from Mr. Chaney's statement whether his weight in air is against brass or other weights; therefore the vacuum reduction above applied is uncertain by a small amount. O. H. TITTMANN.

Washington, D.C., March 10.

MR. TITTMANN thinks the true relation is, without doubt1 pound = 453'5924277 grammes.

Prof. W. H. Miller determined it to be

I pound = 453'59265 grammes, which is the value given on the Card of Equivalents published by the Board of Trade. If the determination quoted by Mr. Tittmann from a Board of Trade Report of 1884 was made under such conditions as to render it authoritative, it is a pity it has been allowed to remain for eight years buried in a Bluebook. One would have expected some intimation of it to be given to scientific men through the Royal Society or in the pages

of NATURE.

As regards the three "standard pounds" which were compared with standards at the Bureau International in 1883, Mr. Tittmann says they had known corrections. This is not stated in the Travaux et Mémoires, where the account of the comparison is given. There is, however, in the case of the two which are of gilded bronze, a reference to a description of them by Prof. Miller in his paper on the standard pound (Phil. Trans. 1856), and, on turning to it, I find that their errors, as stated by him, do not agree even approximately with the determinations made at the Bureau. They differ even in the first significant figure of the error, which is the sixth figure of the entire value; so that, as far as this evidence goes, the five figures 45359 are all that are certain.

a system of three-hourly observations of the dust particles was started, and this has been kept up with but few nterruptions since. The dust observation is made immediately after the usual hourly set is completed, and it can thus be studied along with all the other hourly records in their relation to the prevailing weather.

A great many observations have in this way been accumulated during the past two years, but we have not had time for studying them in detail yet. A mere inspection, however, brings out some interesting points. One of these is the enormous variation that is observed in the number of dust particles, not only in the course of the year, but often in the course of a few hours. At sea-level the number of dust particles in the air at any time depends very much on the locality and on the wind, whether blowing from a polluted district or not. The mean of a number of observations made by Mr. Aitken at Kingairloch, in the west of Scotland, is 1600 particles per cubic centimetre. In London, on the other hand, he found 100,000 per cubic centimetre, and in Paris rather more. On Ben Nevis the mean is 696 per cubic centimetre, the maximum being 14,400 per cubic centimetre, and on several occasions the minimum fell to o. A general mean does not convey a fair idea of the dustiness of the air at the mountain-top, although it may do so for places at sea-level, because there is at the former place a great daily range in the number of dust particles, depending on the rise and fall of the air past the place of observation. If there is any marked variation at sea-level it is entirely of a different character. Below are the means, as well as the maxima and mimima, of all the months that have a fairly representative number of observations.

Number of Dust Particles per cubic centimetre on Ben Nevis.

The above table shows that the Ben Nevis air contains most dust in spring, and it is probable that sea-level air is in this respect similar; the cause of this greater amount of dust in spring than at any other time of the year being the great annual westward motion of the whole atmosphere, or at least of a considerable depth of it, at that time of the year. In a recent paper on "The Winds of Ben Nevis" (Trans. R.S.E., vol. xxxvi. p. 537), it has been shown that this is one of the very few points in which the high- and low-level winds agree, viz. in the excess of easterly winds in spring. The above means for summer are probably too low, as that summer was exceptionally cold, and the general circulation was very abnormal, and that in the direction which would tend to give low dust values. The maximum, 14,400, was observed at I p.m. on April 11, 1891; and, as an instance of how very much the values change in a short time, at 8 a.m. that morning the number was only 350 per cubic centimetre, and by midnight it had again fallen to 600 per cubic centimetre. The daily variation is fairly well indicated from the three-hourly observations. For the months of March, April, and May, 1891, the following are the means for the eight hours of observation :

Here there is a minimum indicated (526) at 4 a.m., and a maximum (1438) at 4 p.m. All the forenoon values are below the mean, and the evening values above it. It would appear that during the forenoon the summit of Ben Nevis is above the first or lowest cloud or dust stratum. About noon this stratum rises to the level of the summit, and during the afternoon hovers above it, and falls again late at night. From this it might be inferred that the summit is oftener clear of cloud in the early morning, and oftener enveloped in the afternoon. A table showing the number of times the top was clear during the last seven years shows that only about 30 per cent. is clear weather in which the summit is free from fog; but it does not show a daily variation as indicated by the dust values, what little it does show being quite the reverse-namely, a maximum of clear weather in the middle of the day and a minimum at night. This clearly indicates that when the dust layer falls below the summit at night, radiation at once forms an independent cap on the hill-top; and again in the afternoon, although the dust stratum envelopes the summit, the opposite radiation warms it up and prevents condensation, or rather evaporates the watery particles of the cloud. So that, contrary to public opinion, the best time to visit the summit for the sake of the "view" is the middle of the day, and not the early morning. During fine settled weather the rise and fall of this cloud stratum can be followed, but in average weather the effect of radiation completely masks it. The effect of solar radiation and nocturnal radiation on dust, as particles and as strata, is a problem that has to be studied and worked out. Very little is definitely known about it at present.

In the study of the nature and motions of clouds the dust observations will be of great value. When a fog envelopes the summit, the number of dust particles observed may vary greatly without any apparent change in the thickness of the fog, but as a rule dry thick fog contains a great amount of dust, while thin wet mist contains very little. It is when a thin drizzling mist envelopes the summit that the lowest values are always obtained, and then there is a distinct difference between the conditions at sea-level and those at the summit, the winds at the latter place differing in direction by 90° or more from the winds below, and sometimes the upper winds are blowing straight out from the centre of a shallow low-pressure area, and the dust that rises with the slight ascending currents of the lower strata is almost entirely filtered out before reaching a height of 4400 feet. One of Mr. Aitken's conclusions may briefly be put as follows: Much wind, little dust; much dust, little wind. That dust seems to accumulate in the quietest places is fully borne out by the Ben Nevis observations. This is true not only horizontally, but also vertically, and it seems probable that this is what chiefly determines the position of cloud strata at all heights. And from this we may infer that the motion of clouds is much slower than that of the general aërial currents; and again, since clouds tend to form between currents, and may have as direction of motion the resultant of the directions of these currents, it follows that as indices to the motions of the upper air the velocity and motion of clouds are very unsatisfactory.

Observations of the apparent haziness of the atmosphere are made whenever it is possible, and the relations between the haziness of the air, the humidity, and the number of dust particles, have been found to be the same as what Mr. Aitken pointed out. Briefly, he found that with a constant humidity the haziness increased or diminished with the number of dust particles, and with a constant number of dust particles the haziness depended on the humidity (at least when the air was within 10 or 15 per cent. of saturation); for with increase of humidity the air became thicker, because apparently condensation begins on the dust particles before the air reaches its point of saturation.

October 17, 1891, when the earthquakes abated, and water returned to some of the wells. The appearance of the sea, as viewed from the land, at first suggested the presence of some "great fish," and columns of "smoke" were seen. Those who visited the spot later (Fig. 2) found black

IG. 2.-Part of a sketch of the submarine eruption near Pantelleria, October 1891. (After Ricco.)

scoriaceous bombs rising to the surface, along a line some kilometre in length, extending north-east and south-west, which might well indicate a submarine fissure, the activity being specially great at certain points. Some of the bombs discharging steam ran hissing over the water with the recoil. Many were still very hot inside, fusing zinc (415 C.), and one was red-hot (in daylight), but below 800 C.. Some pieces were thrown 20 m. in the air, as I gather, not so much by their momentum on reaching the surface as by the explosions occurring there. After the explosions the fragments sank, the material having a sp. gr. of about 2'4. The highest temperature of the water obtained was but 1° C. above that of the surrounding sea. Ricco now questions the trustworthiness of the soundings made at the scene of eruption to a depth of 350 m. without feeling bottom, and he was told that fishermen had previously found but 160 m. of water there. Though some saw bubbles rise to the surface, the gases usually emitted in the case of subaerial eruptions were not detected in the sample of water collected, which Riccò suggests may be due to their having been taken into solution by the water lower down. However, there was a smell" as of gunpowder " at the spot; and the dark, basic, spongy matter of the bombs (previously described), "the only solid material erupted," gives out when heated a sulphurous odour, a fact of which Mr. F. Chapman had previously informed me. The eruption terminated on October 25, and the erupted matter disappeared.

Europe, and an exorbitant price is naturally asked for it. In South Africa the Giraffe is practically extinct, being only still met with in a few isolated localities nearly a thousand miles from Cape Town. In East Africa there are still Giraffes, and in places nearer the sea-board; but here, apparently, there are no means of catching them alive, as the natives do not understand how to do it. Here, however, it is that there appears to be most like lihood of obtaining a fresh supply. This will be an expensive business, but unless some steps are soon taken in the matter it seems that the younger generation of England will grow up without knowing what a living Giraffe is like. Their parents have been more fortunate. From the list given below, it will be seen that there have been 30 individuals of the Giraffe exhibited in the Zoological Society's Gardens since 1836, of which 17 have been born there, and 13 acquired by purchase. Of these 30, one was presented to the Royal Zoological Society of Ireland in 1844, five have been sold at prices varying from £450 to £150, and the remainder have died in the Gardens.

29

Sold Oct. 29, 1853.

Born in the Menagerie, March,, March 29, 1853.

Died May 21, 1872.

Νον. 6, 1866.
Dec. 2, 1859.

Sold May 1, 1863.
Died Dec. 18, 1861.

Nov. 18, 1863.
April 21, 1864.

Mar. 31, 1865. April 3, 1865. April 20, 1865. Sold May 31, 1866. Sept. 14, 1866. Died Nov. 6, 1866. Mar. 17, 1867.

do.

June 20, 1881.

Sept. 12, 1869.

April 27, 1874. May 21, 1878. Jan. 8, 1879. July 9, 1886. Nov. 24, 1891. March 22, 1892.1

GIRAFFES.

THE Zoological Society of London, as our readers, know, have lost their last remaining Giraffe, and, for the first time since 1836, no example of this, one of the most wonderful of living Mammals, is to be seen in the Regent's Park Gardens. Nor does it seem likely that the loss can be easily restored. At the present time, owing to the Mahdists having closed the Soudan to trade, the Giraffe market is very poorly supplied. Only one specimen of this animal, we are told, is for sale in NATURE, vol. xlv p. 251.

NOTES.

THE ordinary general meeting of the Institution of Mechanical Engineers will be held on Thursday evening, May 5, and Friday

The President

evening, May 6, at 25 Great George Street, Westminster, by chair will be taken at half-past seven p m. on each evening, by the President, Dr. William Anderson, F. R. S. will deliver his inaugural address on Thursday evening, after which the following papers will be read and discussed, as far as time permits :-Research Committee on Marine-Engine Trials : Report upon trial of the steamer Ville de Douvres, by Prof. Alexander L. W. Kennedy, F.R.S., Chairman (Thursday, and discussion continued on Friday). On condensation in steam