462 Mr. J. Prestwicti on Tables [ June 18 ; XXII .
" Tables of Temperatures of the Sea at various Depths below the Surface , taken between 1749 and 1868 ; collated and reduced ; with Notes and Sections/ - ' By Joseph PeestwicH ; F.B.S. ; F.Gr .
S. Received June 4 ; 1874 .
( Abstract .
) This paper was commenced by the author more than twenty years since , with a view to the geological bearing of the subject , but was for some years unavoidably interrupted .
It has now been brought down to 1868 , the date of the 4 Lightning ' expedition , when the subject was taken up by Dr. Carpenter , by whom it has since been so ardently and ably carried on .
Nevertheless , as Dr. Carpenter 's work relates almost solely to recent investigations , the author considers that there is yet considerable interest attached to the work of the earlier observers from 1750 to 1868 , though he feels that much of it is necessarily superseded by the great and more exact work subsequent to 1868 .
He is aware that the older observations have also not been deemed reliable on account of the error caused by pressure on the thermometers at depths ; but this is far from applying to the whole of them , as that error was taken into account so early as 1836 , if not before , and a large number of these observations are equally reliable with the more recent ones , while the greater part of the others admit of corrections which render them sufficiently available .
In 1830 , G-ehler gave a list of 226 observations , and D'Urville , in 1833 , tabulated 421 experiments according to depths .
The present paper contains a record of about 1300 observations , which are arranged according to the degrees of latitude:\#151 ; 1st , for the northern hemisphere ; 2nd , the southern hemisphere ; 3rd , inland seas .
They are all reduced to common scales of thermometer , measure of depth , and meridian .
Their position is given on a map of the world , and the bathymetrical isotherms from the Poles to the Equator , based on the correct and corrected observations , are given in a series of ten sections .
The author does not claim for these observations the exact value , or the unity and completeness of plan , of the more recent ones , while , as compared with them , the depths at which they were made are on the whole very limited ; still they include a few at great depths ; and as they extend over much ground that has not been covered by the expeditions of the ' Lightning , ' 4 Porcupine , ' and 4 Challenger , ' he trusts that these Tables may be of some use as comple-mental to these later researches , and as bringing together and reducing to a common standard , observations scattered through a large number of works and memoirs .
At the same time , the author would observe that he thinks it due to our many distinguished foreign colleagues who have been engaged in the inquiry , and whose work seems but little known , that the I'esults of their researches should be understood in this country .
Their conclusions , which are in close agreement with those formed , entirely in1874 .
] Temperatures of the Sea various Depths .
dependency , upon recent and better data by Dr. Carpenter , acquire , from this concordance , additional force and value .
The author was not at all aware himself , in the earlier part of the inquiry , how much had been done , and often found himself framing hypotheses which , on further examination , he found had been long before anticipated by others .
The first part of the paper consists of an " Historical Narrative , " which embraces an account of the character , number , and position of the observations made by Ellis ( 1749 ) , Cook and Forster ( 1772 ) , Phipps ( 1773 ) , Saussure ( 1780 ) , Peron ( 1800 ) , Krusenstern ( 1803 ) , Scoresby ( 1810 and 1822 ) , Kotzebue ( 1815 ) , Wauchope ( 1816 and 1836 ) , John Eoss and Sabine ( 1817 and 1822 ) , Abel ( 1818 ) , Franklin and Buchan ( 1818 ) , Parry ( 1819 , 1821 , 1827 ) , Sabine ( 1822 ) , Kotzebue and Lenz ( 1823 ) , Beechey ( 1825 ) , D'Urville ( 1826 ) , FitzKoy ( 1826 ) , Blossville ( 1827 ) , Graah ( 1828 ) , Berard ( 1830 ) , Vaillant ( 1836 ) , Du Petit Thouars ( 1836 ) , Martins and Bravais ( 1838 ) , Wilkes ( 1839 ) , James Eoss ( 1839 ) , Belcher ( 1843 and 1848 ) , Airne ( 1844 ) , Kellett ( 1845 ) , Spratt ( 1845-1861 ) , Dayman ( 1846 ) , Armstrong ( 1850 ) , Maury , Eogers , Bache ( 1854-57 ) , Pullen ( 1857 ) , Wiillerstorf ( 1857 ) , Kiindson ( 1859 ) , E. Lenz ( 1861 ) , Shortland ( 1868 ) , Chimmo ( 1868 ) .
The second part relates to the " Method and Value of the Observations .
" Wanting a reliable *self-registeriug thermometer , the early observers , for a considerable time , vised a machine contrived by Dr. Hales to bring up water , by means of a bucket with valves , from the depth at which the temperature was to be taken .
This was nsed by Ellis , Cook , Scoresby , Wauchope , and Franklin , and one of a form improved by Parrot was employed by Lenz .
Scoresby 's observations in the seas around Spitzbergen are of much interest .
He showed that while at the surface the temperature varied from about 29 ' to 42 ' , the temperature at depths of from 2000 to 4000 feet was generally about 34 ' to 36 ' ; and there is reason to believe that , with the very slight corrections suggested by Lenz 's subsequent researches , most of them are correct within a fraction of a degree .
The most remarkable readings , however , taken with this apparatus were those obtained by Lenz in Kotzebue 's expedition of 1823 .
He applied to the observations a correction founded on Biot 's law of the variations of temperature experienced by bodies in passing through mediums of different temperature , and determined the lowest temperatures hitherto noted in intertropical seas .
Thus , one sounding in mid-Atlantic , 7 ' 21 ' N. lat. , at a depth of 3435 feet , gave a corrected reading of 35''8 F. , and another at a depth of 5835 feet , in mid-Pacific , 21 ' 14 ' N. lat. , gave 36''4 F. , the surface temperatures being 780,5 and 790-5 .
His observations on the specific gravity of sea-water are also valuable .
, Saussure and Peron used thermometers surrounded with non-conducting substances , so that they might pass through the warmer upper strata of water with little change .
Saussure 's experiments deserve notice , inasmuch as , after applying a correction , they recorded , at that early period , for 461- Mr. J. Prestwick on Tables of [ June 18 , the Mediterranean , at a depth of 1000 to 2000 feet , a temperature , so nearly right , of 55''5 .
Sir John Eoss and Admiral Spratt sometimes used Six 's thermometers , and at others took the temperature of the silt brought up from the bottom .
The former obtained readings of 28'-5 P. for Baffin 's Bay , and the latter of about 55 ' for the Grecian archipelago , agreeing therefore closely with good thermometrical observations .
Phipps used a differential overflow thermometer invented by Cavendish , but it was not found to answer .
This form of instrument remained in abeyance until a greatly improved form of it was contrived by Wal-ferdin ( thermometre a cleversement ) in 1836 .
It was used by Martins and Bravais in the Arctic seas , and by Aime in the Mediterranean , and was said to give very satisfactory results .
Aime also used another somewhat similar instrument , which , at a given depth , was reversed and then hauled up .
These instruments have the great advantage of being free from errors arising from the shifting or immobility of the index .
It is not clear why their use was abandoned , except that they were difficult to construct and not generally known .
Six described his thermometer in 1782 ; but the first person to use it was lvru sen stern , in 1803 .
It did not come into general use for deep-sea observations until the Arctic voyages of Eoss and Parry , after which date it was , with the exception of Lenz 's and Aime 's , employed for that purpose on all the expeditions sent out by foreign governments , as well as by our own .
The necessity of protecting the instrument against pressure was early insisted upon by Lenz , Arago , Biot , and others ; and there is reason to believe that protected thermometers were used by D'Urville and Berard , for their observations in the same Mediterranean area show a remarkably close agreement with those recently made by Dr. Carpenter , with protected instruments , at and below depths of about 200 fathoms , the results being:\#151 ; D'Urville ( May 1826 ) .
Berard ( Nov. 1830 ) .
Carpenter ( Aug. 1870 ) .
Surface-----64'-1 P. Surface_____67'* 1P .
Surface____73'-5 P. 1062 ft ... ..54'-2 3189 ft ... ..55'-4 2958 ft ... ..55'*5 3189 ft ... ..54'-7 6377 ft ... ..55'-4 7968 ft ... ..54'-7 It was , however , on Du Petit Thouars 's voyage of 1836 that the first special steps were taken to protect the thermometer against pressure , lor that purpose an improved instrument of Bunten 's was provided , and this was enclosed in a strong brass cylinder .
Pifty-nine observations were made , of which Arago reported that 21 might be considered perfectly good .
Temperatures of 36 ' , 37 ' , and 38 ' P. were recorded at depths ( 900 to 1100 brcissdtS ) in both the mid-Atlantic and mid-Pacific ; while in one case , in taking a sounding at a depth of 12,271 feet near the equator in the Pacific , the instrument came up crushed , but with the iudex fixed at 34'-8 P. ( l'-6 or l0,7 C. ) , In a certain number of cases ( 24 ) the 1874.1 Temperatures of the Sea at various Depths .
pressure forced water into the cylinder .
For these corrections were made . . .
, .
In 1839 , MM .
Martins and Bravais made a series of observations in the sea between Norway and Spitzbergen with instruments carefully protected against pressure by means of glass tubes or metal cylinders .
They used both self-registering thermometers ( thermometrographes ) and Walferdin 's self-registering overflow thermometers , sending down two to four of each in every sounding , and taking the mean of the readings .
These probably are amongst the most accurate observations on record .
To a great extent they confirm those of Scoresby ; and they further showed that the bottom-temperature near the Spitzbergen glaciers was about 29 ' F. None of the soundings exceeded 3000 feet .
In 1857 , the late Admiral FitzBoy furnished Captain Pullen with thermometers specially constructed to resist pressure , and some very interesting , though somewhat variable results , were obtained therewith .
On two occasions a temperature of 35 ' F. was recorded\#151 ; one in the Atlantic , 26 ' 46 ' S. , at a depth of 16,200 feet , and the other in the Indian Ocean , at a depth of 13,980 feet .
With regard to the many observations made with unprotected instruments , they mostly admit of correction , which renders them available .
Such corrections have been independently computed , with little difference , by Du Petit Thouars , Martins , Aim , and the late Dr. Miller .
The author , taking the mean of their estimates , uses as a coefficient \#151 ; 1 ' F. for every 1700 feet of depth .
In the third part of the paper the author shows the " State of the Question at the date of the Lightning Expedition .
" Ellis , Forster , Peron , and others early remarked on the decrease of temperature at depths in temperate and tropical seas , but it was not until 1823 that Lenz showed that a temperature of 35 ' to 36 ' existed at greater depths in those seas .
Notwithstanding this , D'Urville in 1826 , misled by incorrect readings obtained by previous observers with uncorrected instruments , and in the absence of sufficiently deep observations of his own , was led to believe that the temperature in open seas at and below a depth of 3214 feet ( 600 brasses ) was nearly uniform at 39'*8 F. ( 4'*4 C. ) , and that between the latitudes of 40 ' and 60 ' there is a belt of a like nearly uniform temperature .
A few years later , Arago , discussing the results obtained by Du Petit Thouars , insisted that they effectually disproved this hypothesis .
Nevertheless , in 1839 , Sir James Boss made the same mistake as D'Urville , and unfortunately obtained for it a wider circulation , which seems , however , to have been almost altogether restricted to this country .
" Still , Boss 's numerous observations , when viewed under correction , are of considerable value , though the author considers that some error has occasionally crept into that uniform reading , so often recorded , of exactly 390,5 .
Both D'Urville and Boss wrote under the opinion that sea-water , like fresh water , attained its maximum density at a tempera466 Mr. J. Prestwich on Tables of [ June 18 , ture of between 39 ' and 40 ' , \#151 ; a point that had been investigated and disproved by Marcet in 1819 , approximately determined by Ermann in 1822 , and which was finally settled by Despretz , in 1837 , at 25''4 E. While the law of the decrease of temperature with the depth , in both the great oceans , to a point but little above the zero of Centigrade was being established , experiments had been carried on in polar seas showing , on the contrary , that the temperature at depths was higher than the average surface-temperature .
The careful experiments of Scoresby and of Martins fully established this for the Arctic seas , and those of Ross , after correction , establish the same fact for the Antarctic Ocean .
In one part , however , of the Arctic seas this rule has not been found to hold good ; for , in Baffin 's Bay , the experiments of John Ross , Sabine , and Parry , at depths of from 600 to 6000 feet , agree in showing a decrease of temperature of from 30 ' to 32 ' near the surface , to 29 ' and 280,5 at the greatest depths attained .
There are also two instances given of yet lower temperatures .
Nor were observations wanting in inland seas .
Those of Saussure , D'Urville , and Berard had indicated generally that , in the Mediterranean , the temperature decreased to a depth of about 1200 feet , after which it remained uniform at from54 ' to 55'E .
; and , in 1844 , Aim instituted a series of experiments which resulted in showing that the diurnal influence ceased to be sensible at a depth of from 16 to 18 metres , and the annual variation at a depth of from 300 to 400 metres , below which the temperature remained constant at 12'-6 C. ( 54'-6 ) ; and this he showed to be the mean winter temperature of the area of the Mediterranean , over which his observations extended .
These observations were confirmed , for the Eastern Mediterranean , by those of Admiral Spratt .
His first experiments in the Grecian archipelago showed , at a depth of 1200 feet , a temperature of 54'-5 to 55 ' E. , while the later ones , at greater depths in the open sea , give , after correction , a temperature of about 55 ' .
In the Red Sea , Captain Pullen found that while the surface-temperature varied from 77 ' to 86 ' E. , it fell to 70 ' or 71 ' E. at 1200 to 1400 feet , below which it remained uniformly the same to the greatest depth he attained of 4068 feet .
Some curious results were obtained in 1803-6 by Dr. Homer in the Sea of Okhotsh .
The surface-temperature was 46'-4 E. ; and the author finds ( after correcting the original readings ) at 360 feet a temperature of 28 ' , and at 690 feet of 28''6 , which is almost exactly that determined by Despretz as the temperature of sea-water at the moment of congelation .
The cause of the decrease of temperature with the depth in the great oceans was early investigated by physicists .
Humboldt concluded that " the existence of these cold layers in-low latitudes proves the existence of an undercurrent flowing from the poles to the equator .
" D'Aubuisson and Pouillet took the same view .
D'Urville went further , and remarked that " it is rather a transport nearly in mass , and very slow , of the deep waters of high latitudes towards the equator , " and that from his zone of 1874 .
] Temperatures of the Sea at various Depths .
40 ' to 60 ' lat. there are two insensible currents\#151 ; a lower one towards the equator , and an upper one towards the poles .
Arago saw no other explanation than " the existence of submarine currents carrying to the equator the bottom waters of the icy seas .
" We are , however , indebted to Lenz for a full and philosophical review of the whole subject in 1847 .
After showing that all the facts proved the existence of a temperature of from 34 ' to 35 ' F. at depths in the tropical seas , and that this could only be maintained by a constant slow undercurrent from the poles to the equator ( which , on the other hand , must necessitate the transfer by an upper current of the equatorial waters to the poles ) , he proceeds to show by a series of observations , chiefly those of Kotzebue , and by a diagram , that a belt of cooler water existed at the equator , and that the temperature , at equal depths , was lower there than a few degrees to the north and south of it ; and he concluded that this arose from the circumstance that the deep-seated polar waters there met and rose to the surface .
As corroborating this view , he showed that the waters in the same zone were of lower specific gravity , a fact that had been before noticed by Humboldt .
The author then proceeds to consider some " G-eneral Conclusions .
" Some of these have now been better established by the more recent expeditions and by the researches of Dr. Carpenter .
Taking , however , other areas , he shows that in the Arctic Ocean the bathymetrical isotherm of 35 ' is deepest on the west of Spitzbergen , while nearer Greenland and again nearer Norway the deep waters are colder .
The several isothermal planes of 40 ' , 50 ' , 60 ' , 70 ' , and 80 ' are then traced southward , attainmg their maximum depth between 50 ' and 40 ' lat. , and rising thence towards the equator .
Section No. 2 , from Baffin 's Bay to the equator , shows that the higher isotherms are not prolonged so far north as on the first line , and that the water at the bottom of the bay is colder than in the Spitzbergen seas , approaching much nearer that of its maximum density and of its point of congelation ; whence he concludes that this is the main source of supply of the deep-seated cold waters in the Atlantic , which , after attaining their greatest depths between latitudes 40 ' to 50 ' IN .
, are found 3000 to 4000 feet nearer the surface on approaching the equator .
In the South Atlantic , the bathymetrical isotherms show lesser curves ; and while the isotherm of 40 ' crops out between the lat. of 50 ' and 55 ' , that of 35 ' is prolonged into high southern latitudes on a nearly uniform plane of .
7000 to 8000 feet deep .
In the Pacific , the sections show that , notwithstanding there is no appreciable polar current through Behring 's Straits , the bathymetrical isotherms of 60 ' , 50 ' , and 45 ' do not extend so far north as in the Atlantic , while that of 35 ' is apparently not prolonged beyond 60 ' N. lat. As the presence of temperatures lower than those which prevail in On Tables of Temperatures of the Sea .
[ June 18 , parallel latitudes in the Atlantic cannot be due to north polar waters , a and seems more than could be maintained by local influences , the author o concludes that the effect may probably be due to waters from the Antarctic 3 Ocean , of the presence of which the low temperatures at depths throughout rj the Pacific affords evidence , passing , in the absence of any counter flow , to the extremity of the North Pacific , where they are thrown upwards by the rising slopes of the ocean-bed , as on banks in open oceans .
On the other hand , in the South Pacific the conditions seem very similar to those in the South Atlantic .
The bathymetrical isotherms appear , however , to be prolonged further south than in the South Atlantic , which arises pos-sibly from the circumstance that as none , or comparatively none , of the warm equatorial water can pass into the Arctic Ocean , a larger proportion passes into the Antarctic seas .
In the Southern and Indian Oceans the conditions seem analogous to those of the North Pacific , only they are more masked by the high surface-* I temperatures of the Arabian Gulf .
The author agrees in the opinion which has been advanced of the flow over the ocean-bottom of cold undercurrents at and below 35 ' , one from the north and the other from the south pole to the equator , and of their rise in the equatorial regions of the Atlantic .
They must , then , necessarily tend to disperse and escape into other areas ; but whether by a movement in mass of the upper strata , or by currents in more definite channels , or by both causes combined , remains to be proved by further research .
He inclines to the latter view .
He would suggest the question whether the Gulf-stream , together with others which seem to originate or acquire additional power in equatorial seas , such as the Guinea and Brazilian currents , may not receive either their initial start or be strengthened and maintained by the surging-up of the Arctic and Antarctic waters at the equator , while another portion of those waters may be deflected back in insensible currents to polar regions .
In the same way some of the great currents of the North Pacific may arise .
The paper concludes by a review of the other causes connected with these conditions , by a consideration of the normal isotherms of the polar regions , and by a comparison of the temperatures of inland seas , which are dependent on local climatal conditions , with those of the great oceans , which are subject to such vast distant influences ; and he directs attention to the important bearing which these questions of oceanic physics have on many geological problems .

