May 8, 1919] 



NATURE 



industries. We can keep this increased trade only if 

 we maintain West Indian production and, what is 

 quite as urgent, improve West Indian grades so that 

 they can compete with the Mediterranean. This may 

 or may noti be achieved by means of sponge culture, 

 but it is worth trying. The Americans have u'nl 

 doubtedly made progress with sponge culture in 

 Florida, and a significant fact is recorded in a recent 

 British Colonial Report on the Turks and Caicos 

 Islands to the effect that at one of these islands 

 8000 acres of sea for sponge culture has been conceded 

 to a capitalist from New York, While we should 

 prefer to see British enterprise of this nature, par- 

 ticularly in a British Possession, we have to recognise 

 a certain consistency in United States action. Most 

 of the marine investigation in the West Atlantic has 

 been American ; for instance. Prof. Nutting's recent 

 and former expeditions, the studv vears ago on the 

 fishes of Porto Rico by the U.S. "Government, and 

 the quite recent oceanographic work in the steamer 

 Bache. It is to be hoped that Great Britain will see 

 ■its way to take up the sponge question, first from the 

 scientific, and then from the commercial, point of 

 view, and that a start will be made at the earliest 

 possible date. W. R. Dunlop. 



Seaholme, Hythe, Kent, April 23. 



Wasps. 



The warm spring weather which made its advent 

 on Good Friday (April 18), and was continued on 

 following days, brought out numbers of humble-bees, 

 a few wasps, and butterflies of various kinds. I have 

 usually observed that the humble-bees precede the 

 wasps by a week or two. 



A wasps' nest (Vespa germanica) situated in the 

 garden here in 1915 was a rather strong one, and 

 on digging it out in October I estimated the number 

 of cells as 12,900. A nest of the same species which 

 I had in 1918 was much stronger. In 1915 the hourly 

 number of wasps flying in and out of their nest was 

 6500 at the most abundant period, while in 19 18 the 

 rate was no fewer than 15,500. The record heavy rains 

 of September last, however, swamped the nest and 

 brought it to a premature termination, when but few 

 of the young queens had taken to flight. If the nest of 

 19 18 had a number of cells proportionate to that of 

 19 15, according to the hourly rate of wasps flying to 

 and fro, then the total number of cells must have 

 been about 30,000; but I prefer to take a more 

 moderate estimate, and to put the aggregate at 25,000. 

 I could not, however, actually determine the number 

 by observation, the layers of comb being so soaked 

 with the wet that they did not admit of detailed in- 

 vestigation. If each cell produces three generations 

 of wasps, then my nest of 19 18 must have been 

 responsible for quite 75,000 wasps. Needless to relate, 

 house-flies were not troublesome in this neighbour- 

 hood during last summer! But which pest of the 

 two, wasps or house-flies, is the more tolerable? For 

 my part, I greatly prefer the wasps ! 



Can any reader inform me as to the number of 

 wasps supposed to be associated with a very strong 

 nest? W. F. Denning. 



Bristol. 



185 



THE LUNAR TIDE IN THE ATMOSPHERE. 



TIDAL theory was first applied with any 

 success to the atmosphere by Laplace, and 

 he also first attempted to determine the tidal varia- 

 tion of pressure from barometric observations. 

 His material consisted of 4752 measurements of 

 the height of the mercury column at Brest (lat. 

 49° N.). These were far too few for the purpose, 

 NO. 2584, VOL. 103] 



however, and his result, given in tome v. of the 



j "M^canique Cdeste, " cannot be regarded as a 



I determination of the quantity .sought for, which is 



much smaller than Laplace's value. Another lunar 



reduction of barometric data from Brest was made 



about thirty years ago by Bouquet de la Grye, 



i but his series of observations (consisting of hourly 



' values extending over a few years), while larger 



i than that used by Laplace, still seems to have 



I been inadequate. He arrived at a lunar daily 



inequality of pressure which was not by any means 



nearly semidiurnal in type, though the semidiurnal 



component — 0*020 sin {2t+ioo°) mm. of mercury 



— was larger than the probable true value of the 



tidal variation at Brest. 



The atmospheric tide was determined from a 

 tropical series of barometric records so early as 

 1847. There now exist more or less trustworthy 

 determinations for five tropical stations — St. 

 Helena, Singapore, Samoa, Hong-Kong, and 

 Batavia. The results for the two last are from 

 long series of hourly observations, extending over 

 thirty or more years, and are therefore of con- 

 siderable accuracy. Though the tidal barometric 

 variation has its maximum value at the equator, 

 its magnitude there is very small. At Batavia 

 (6° S.) it may be represented by the formula 

 0*065 sin (2^4- 65°) mm. of mercury, 



where t denotes time reckoned from lunar transit 

 at the rate of 360° per lunar day. The phase 

 angle 65° indicates that maximum pressure occurs 

 nearly an hour after the moon crosses the meri- 

 dian. 



Until recently the only determination of the 

 tide which could be considered as probably an 

 approximately true one, among the results for 

 extra-tropical stations, seems to be that obtained 

 by Morano from five years' hourly barometric 

 observations at Rome (42° N.). Though the 

 series of data was not large, the resulting ampli- 

 tude and phase agree with what might be ex- 

 pected in this latitude. Many other attempts to 

 determine the tidal barometric variation in Euro- 

 pean latitudes have been made without success. 

 The most important of these investigations was 

 due to Airy, who dealt with as many as 160,000 

 hourly observations of the barometer at Green- 

 wich (51° N.), ranging over the twenty years 



1854-73- 



The barometric pressure is affected by a solar 

 semidiurnal variation as well as, and of much 

 greater amplitude than, the lunar tidal variation. 

 Unless the former is properly abstracted from the 

 hourly values before deducing from them the lunar 

 inequality, the determination of the latter may 

 be seriously affected by a residuum of the solar 

 term. Two other causes operate to enhance the 

 difficulty of detecting the lunar variation in the 

 barometric records of stations in moderate and 

 high latitudes. The first is the rapid diminution 

 of the tidal amplitude as the latitude \ increases. 

 The second is the increase in the irregular fluctua- 

 tions of the pressure. At Brest or Greenwich 

 these range over several millimetres (of the mer- 



