Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. 182 Cadell on the Navigation of the Murray . By the same mail which takes this hurried communication (as I leave to-morrow on my second voyage) I send newspapers and a pamphlet containing articles on the river, which I trust may be found of interest. Since my return, the Legislative Council of this colony have awarded to me a "gold medal" for opening up and showing the capabilities of the river, and have also voted me 4000/. as a bonus, to place two steamers on the upper rivers. Next flood I look for- ward to having five steamers on the river, and a boat to go out- side. These steamers, with a sufficient number of barges, will be amply sufficient for all goods and passenger traffic. Indeed, the Governments of all the colonies, as well as the settlers on the river, have given me every inducement to prosecute the enter- prise. When the new steamers are on the river I look forward to reaching the Mitta Mitta by the Home ; Seymour, about 40 miles from Melbourne, by the Goldburn, between Yass and Gundagai on the Murrumbidgi, and Wangaratta on the Ovens : as to the Dar- ling, I am afraid to hazard an opinion. I did myself the honour of addressing you by letter, dated " Lady Augusta, 5th Oct., 1853," proffering my services to leave any stores on the Murray or Darling, which might possibly be re- quired for the " Exploring Expedition " which, under your auspices, is to be started for the interior. I can now only repeat what I said then. My time, since my return to Adelaide from the Murray, has necessarily been much occupied in carrying out affairs connected with a navigation of such recognised importance, which I trust will serve as an excuse for the hurried manner in which these remarks have been thrown together ; and, looking forward some day to be of substantial use to your learned body, I have the honour to be, Sir, Your obedient servant, Thomas Cadell. XL — Notes on the Passage of Hannibal across the Alps ; and on the Valley of Beaufort, in Upper Savoy. By Professor Paul Chaix, of Geneva, Corresp. f.r.g.s. Read, May 24, 1855. That " vexata qusestio," the passage of Hannibal, has recently undergone fresh investigation in a work by Mr. Schaub, of Geneva. The author begins with a brief summary of opinions of certain previous writers. Many of the views that have been advocated Chaix's Notes on the Passage of Hannibal across the Alps. 183 appear to have been suggested by a desire to identify the line of march with the neighbourhood in which the inquirers resided, or by similar circumstances of an accidental nature. Mr. Schaub regards the theory of Whittaker, followed by Simler and Grosley, which leads Hannibal over the Great St. Bernard, as the most paradoxical. Deluc and Melville produced the most remarkable work on the subject, in three editions, the first of which appeared in 1818. It was ably supported by Wickham and Cramer in 1820, who in- creased its importance by a few additions and modifications These authors adopt Roquemaure as the point where Hannibal crossed the Rhone, and Vienne as the northern extremity of his march up the left bank of the river. The triangular interval between the Rhone and the Isere in Dauphiny, compared by Polybius to the Egyptian Delta, answers to the Insula Allobrogum. The Cartha- ginian General entered Savoy at St. Genix d'Aosta, and crossed the hilly country at Mont du Chat (Mons Thuates), where his first encounter took place with the mountaineers, who kept watch during the day only. From thence Hannibal advanced through the fine plain of Chambery, and along the wide and fertile valley watered by the Isere, to Bourg St. Maurice, crossing the Alps at the pass of the Lesser St. Bernard, where he fought his last battle with the mountaineers, and then halted on its level plateau for two days, waiting for stragglers. M. Deluc attached too much importance perhaps to the dis- covery of a great white rock and of large bones upon the Lesser St. Bernard, and of a silver shield which was found by a farmer in 1714, and deposited in one of the Parisian museums. The shield was dug out of the ground on the estate Du Passage, between La Tour-du-Pin and Abrets, half a mile from the road between Vienne and Chambery. It was ornamented with Punic intaglios and bas-reliefs, and considered to have been a votive offering. The best arguments of Deluc, Melville, Wickham, and Cramer were based on the concordance of their itinerary with the distances, the number of days, and the geographical data men- tioned by Polybius ; also on the adoption of that text, almost exclu- sively, as the only faithful guide ; and further, on the coincidence of that itinerary with a choice of locality and season such as would be made by a man of common sense, and particularly by a general commanding an army. A different opinion was maintained by the Marquis of Saint Simon, who attempted to prove that the Cartha- ginians marched along the river Ubaye to Mount Viso. This route- leaves nine days to be accounted for. The author supposes that the army was misled by the guides, and that the time was spent in ascending the mountain. This theory affords the author 184 Prof. Paul Chaix's Notes on the an opportunity of finding a pulpit, at a height of 13,000 feet above the sea, for the celebrated speech reported by Livy ! Apparently for the sake of that result, he does not hesitate to trace the route of an army over a lofty mountain, instead of through passes 6000 or 7000 feet lower. But these guides had been chosen to meet Hannibal by the Insubres, a nation in frequent intercourse with Transalpine Gaul, their original country. The Insubres had found their way to Carthago Nova, in Spain, and the guides they had selected were not likely to mistake the safest road to Italy. Han- nibal was in alliance with the Allobroges, as well as with the Insubres, and a prince of the Cisalpine Gauls was with the guides. That road was undoubtedly the safest for him which joined the territories of those nations. In 1851 M. Replat, a barrister of Annecy, published a work entitled ' A Note on Hannibal's Passage,' supporting a theory originated by M. Blanc of Beaufort, and Count Vignet. M. Re- plat leads Hannibal from the Rhone on reaching the Isere, not only below Vienne, but without entering the territory of the Allo- broges, in opposition to the positive evidence of Polybius and Livy. He places the Insula Allobrogum between the Isere and the Drome, although the high and very bold mountains of Royans, Trieves, and Vercors give that tract no resemblance to the Egyp- tian Delta. The " Insula " lies so evidently north of the Isere, that another commentator, M. Latronne, makes the Carthaginians turn back and recross the Isere in favour of his theory, which continues the line of march along the rivers Drac and Durance to the Pass of Mont Genevre. M. Replat, however, follows the left bank of the Isere to Pont- charra, near Fort Barraux, where he places the site of the first battle with the mountaineers, who left the passage defenceless at night. He then crosses to the opposite bank, and reaches Albert- vine and Conflans. From the last-named place MM. Vignet, Blanc, and Replat advocate the claims of a new route, never before suggested. Turning from the great valley of the Isere because it appears to be " too crooked," they endeavour to trace the African General's march along the hitherto unheeded valley of Beaufort and through a labyrinth of defiles to the pass of La Seigne and the Allee Blanche, at the southern base of Mont Blanc. This route is chosen by M. Replat as the shortest way across the great ridge of the Alps by an Eastern pass. Valley of Beaufort. — The Valley of Beaufort lies N.E. of Albertville, between that town and Chamounix. There are four routes between it and the Chapieu, which forms the approach to the western foot of the Col de la Seigne : — 1st. The Pass of La Sauce, which is so forbidding that it would be madness to think of Passage of Hannibal across the Alps, Sfc. 185 it. 2nd. The Pass of Cormet, or Platte, or Croix de Biollay, which presents an accessible tract of pasture-ground, after an ele- vation has been attained of 6441 feet, almost the height of the Lesser St. Bernard. 3rd. The Col du Mont Joly, which I found to be higher. 4th. The Col de la Fenetre, still higher. The 3rd and 4th do not lead directly to the Col de la Seigne, but into the Vale of Montjoie, and then over another pass, the Col du Bon- homme, to the Chapieu, where all these routes converge, leaving the Pass of la Seigne still in advance. M. Replat prefers the 2nd route through the Pass of Cormet, or Platte : but he thinks it not altogether improbable that Han- nibal, " misled by his guides," ascended in succession the Col du Mont Joly, the Bonhomme, and lastly the Col de la Seigne ; three difficult passes, in preference to the single and easy Pass of the Lesser St. Bernard. M. Replat's / Note ' has been answered in the work by Mr. Schaub, mentioned at the beginning of this paper. Mr. Schaub is intimately acquainted by foot journeys with all the passes in the Alps. His thorough knowledge of Greek has made Trim familiar with the chief authority in the original. He follows the names, distances, and times given by Polybius ; he does not dispute the competency of the guides, and he concludes that the Carthaginian General, in choosing his road, kept in view the sustenance and safety of his army, as well as the rapidity of his march. At Pontcharra, the position selected by M. Replat for the first encounter with the mountaineers, Mr. Schaub shows that there is not in the whole neighbourhood any mountain pass answering to the historian's description ; and the distances from the Isere to Pontcharra are also quite inconsistent with Polybius. The Valley of Beaufort is easy of access at its entrance only ; but it soon merges into narrow defiles, with stupendous ridges and precipices, leading to a chaos of mountain passes, at a great ele- vation. On the other hand, the Valley of the Isere is wide and open throughout, from Albertville to the foot of the Lesser St. Bernard. It is almost level, fertile, and now occupied by numerous wealthy villages and towns. M. Replat's supposition that it was marshy and overflowed is quite groundless. The Lesser St. Bernard is one of the easiest passes in the Alps, and lies at the southern base of the Cramont, a mountain identified with the Cremonis Jugum, mentioned by Cselius Antipater as the locality of Hannibal's passage. There is also a table-land on the top of the pass, 6 miles long, and 900 feet lower than the Col de la Seigne, sufficient, it is presumed, for the encampment of Han- nibal's army while waiting two days for stragglers. Three days more were spent in the valley below, owing to the passage being 186 Prof. Paul Chaix 's Notes on the obstructed by snow. The Col de la Seigne, on the other hand, passes the northern base of the Cramont. There is only a narrow ridge on its summit, and there is no flat spot lower down, as M. Replat gratuitously imagines. It is also 900 feet higher than the level plateau of the Lesser St. Bernard. The distance from Albert- ville to the Italian valley of Aosta, along the Isere and over the Lesser St. Bernard, is 50 miles, measured along the road,, which is flat for 34 miles, and the pass is easily crossed in the remaining 16 miles. Following M. Replat's itinerary through the Valley of Beaufort, the distance between the same points is 42 miles, scarcely 8 of which may be termed easy, while the remainder includes two and even three very high mountain passes, in a very rugged, unculti- vated, and inhospitable country. It is also well known that these passes are impracticable at the end of October, when Hannibal's passage occurred ; and as each pass would be a day's march for an army, the saving of 8 miles on level ground appears to be an in- sufficient reason for encountering such obstacles.* As the Valley of Beaufort is very little known, a sketch of its present state may be useful to tourists, and perhaps afford some additional material for Mr. Murray's excellent Guides. The situation of the valley has been already described in a previous page of this paper. It includes within its branches, and the surrounding ridges and passes, an area of 1 65 square miles within a circumference of 48 miles. The chief town, St. Maxim e de Beaufort, is 12 miles N.E. by E. of Albertville, and it is situated at the junction of five valleys, which form together the Beaufort country. The Vale of Hauteluce is on the N.E. ; the Vale of La Gite is on the E. ; Pontcellamot, or Poncelamont, is on the S. ; Trecols and Roselein are on the S.E. ; and the Valley of Beaufort, in which these unite, is on the W. Each valley feeds a stream. The Doron flows from La Gite, and receives the Argentine from Pontcellamot ; the Dorinet, or lesser Doron, from Hauteluce ; and the Trecol and the Roselein from the valleys of the same names. The Doron then pursues its course for 10 miles below Beaufort, till it unites with the Arly, a tributary of the Isere, at Conflans. An indifferent road ascends from Albertville, along the banks of the Doron, to St. Maxime de Beaufort. In other directions * The Dissertation by Count de Fortia d'Urban contains a list of many works on Hannibal's Passage. In 1854, " a Treatise," by Robert Ellis, b.d., was printed at the University Press, Cambridge, and answered in " a Criticism," by W. J. Law, a.m., London, 1855. Mr. Law notices the announcement of M. Chaix's Paper in a postscript, the remarks in which appear to be based on an erroneous assumption of M. Chaix's views.— [See also a work on the same by H. L. Long, 1831. — Ed.] Passage of Hannibal across the Alps, Sfc. 187 the country is only accessible through a dozen cols or mountain paths, practicable only for mules and pedestrians. These passes are, on an average, about 6500 feet above the sea, and the follow- ing are the most remarkable : — On the S.W., La Batia, leading from Areche to La Batia, in the valley of the I sere. On the S., the Louse (? Luce) and Grand Cor met, crossing from the head of the Vale of Pontcellamot to the Isere in Tarentaise. On the N.W., the Col de la Leizette passes between Hauteluce and Flumet, through Notre Dame de Bellecombe. On the N.E., the pass of La Fenetre and the Col Joly lead to Chamounix, through the Vale of Montjoie. On the E., the pass of La Platte, or Croix de Biollay, leads from La Gite, and that of La Sauce from Rose- lein to the valley of the Chapieu in Tarentaise. A few other passes form internal communications from one valley to another such as the Char de Montagne, or Col de Boudin, between Areche and Trecols, and the Plan de FEstace, between La Gite and the upper part of Hauteluce. La Sauce is perhaps the most difficult, especially at the narrow part called Gorge des Caves, but none of these paths is dangerous at a proper season. The time required to reach Moutiers on the Isere from St. Maxime de Beaufort, either through the Grand Cormet or the Col de la Luce, is equal to a day's march. It is 4 hours from Beaufort to the Chapieu over the Col de la Platte ; and half a day from the upper part of Hauteluce to La Gite through the Plan de l'Estace. A difficult march of 1^ hour is required to reach the crest of Mont Joly from the Col Joly, and the descent to Notre Dame de la Gorge occupies the same time ; from thence it is 2£ hours to Contamines. The distance from Flumet to the Col de la Leizette is equal to 4| hours, and the road lies through cultivated land and swampy pastures. Thence to Hauteluce is 1 hour more. None of the mountains in the Beaufort country is of a great height ; but their majestic summits are capped with snow before the autumn begins, especially the Needle of the Grand Fond and the Crest du Rey (King's Summit). Mont Joly, the Point of Arotte, and Rousselette, to the E., rise in front of Mont Blanc. The Rocks of Enclaves are perhaps the most curious feature in the orography of this district. They form a group of mountains in its centre, between the Doron and the Dorinet. An oval, solitary basin is hollowed out in their midst, and surrounded on all sides by granitic ridges so completely that its waters can only be discharged through subterranean channels, and over numerous and magnificent falls, into the wild valley of La Gi'te. A cluster of chalets, named Outray, occupies the centre of this secluded basin. I found there plenty of rock-crystals and samples of copper and iron ore, besides anthracite coal and talcose schists. 188 Prof. Paul Chaix's Notes on the The following is a List of Measured Altitudes, with the Names of the Observers. Positions. French Metres. English Feet. Observers. Albertville 9 > Aixme > » • Lesser St. Bernard, Pass , , Hospital . . J i 3 > .... Chapieu > > St. Maxime de Beaufort 3 > 3 > > > J J , , , , at the Inn, 3) observations J St. Maxime de Beaufort at the Church, ) on a hill J Pass of La Platte, or Croix de Biollay > > » » Pass of La Luce Grand Mont, north of La Luce Pass of the Grand Cormet Pontcellamont, chapel of St. Gueriii 3 3 > f Char de Montague, or Col de Boudin Koselein, or Iioselen > > j > Praz, village S. of Beaufort Areche, church Fontaine, hamlet E. of Beaufort La Glle, at Cyril Freyzon's Cot Pass of Plan de l'Estace, above La) Gite J Lake of La Girotta Pass of Mont J oly it > > , , , , at Chalet Avocat Annuit, in vale of Hauteluce Hauteluce , , (better observations) La Leizette, pass above Hauteluce . . , , (inferior observations) . . Bellecombe, Notre Dame de tt 1 1 .... Flumet, on a slope , , in another part Col de la Seigne Oratoire du Glacier Col des Tours 315-7 338*0 758-0 711*0 2206-0 2151-0 2172*0 1549-0 1546-0 799-1 797*0 754-0 744-1 749-8 1967*0 1975-0 2138*0 2720-0 2139-6 1523-3 1529-0 1730-0 1481-0 1464-0 999-8 1027*7 1033-4 1658*6 2286-9 1712-9 1975-5 2028-0 1632-4 1233-7 1141-7 1158-7 1775-7 1782-8 1118-1 1115-3 897-0 922-5 2519-0 1774-0 2661 '0 2682-0 1033-8 1099-3 2463-5 2310-7 7169-5 6990-7 7059-0 5034-2 5024-5 2617-0 2590-2 2450-5 2436*9 2455-5 6392-7 6418*9 6946*0 8840-0 7007-0 4950-7 4969-2 5622-5 4813-2 4758-0 2694-3 3323-3 3358-5 5390-4 7432-4 5566-9 6420-3 6590-0 4470-9 4009*5 3710-5 3765-7 5770-0 5794-1 3633-9 3624-7 2915-2 2998-1 8186*5 57655 8648-3 8716-5 A. Beaumont. Billiet. Beaumont. Billiet and Gravier. Escher. Billiet. Billiet and Gravier. Studer. Favre. Beaumont. Beaumont. Favre. Chaix. Chaix. Studer. Favre. Favre. Favre. Favre. Favre. Favre. Favre. Favre. Favre. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Favre. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Chaix. Favre. The geological structure of the Beaufort country offers the following indications : — Mica-slate forms the slope of the moun- Passage of Hannibal across the Alps, fyc. 189 tains from Flumet up to La Leizette ; but the crest of the pass is clay slate, dipping to the S.E., and it is also seen along all the crests as far E. as the Col Joly. It is found again at the Cormet, dipping N. ; but on the southern slope of that pass it dips S. Clay- slate rises in immense masses between Areche, Petit Cceur, and Cevins, where abundant quarries are worked, yielding very large fine-grained and shining slabs. Mines of anthracite coal are worked near Petit Coeur, and others have been recently opened at Areche in the Vale of Pontcellamot. Mountains of granitic slate form the eastern slope of Pontcellamot, and occur again in the vale of Trecols and on the southern slope of the Cormet. The Doron flows for miles in the vale of La Gite, through a narrow and deep cleft in talcose schist, dipping to the S. and E. Each of the smaller valleys has peculiar features. The vale of La Gite is without doubt the finest. At its head a waterfall, issuing from the wild gorge called Les Caves, in the pass of La Sauce, unites with another stream in a flat meadow to form the torrent of Doron. From thence the Doron plunges into a deep ravine, extending westward between vertical walls of rock, and becomes an uninterrupted cascade for many miles, increased by numerous streams which descend in foaming sheets from the rocks of Enclaves on its right bank, while the left bank is shaded by tall and dark pines. Scenery equally wild and magnificent forms the entrance from this valley into Roselein. The vale of Pontcellamot in its' lower part is more open, and the forests which crown its sides surround a basin of cultivated land and slopes of smiling pastures. The number of hamlets bespeak a larger and more wealthy population scattered around the larger village of Areche. A pretty fall is formed by the river Argentine at Areche. A winding path ascends to the second terrace of this valley, where the forests encroach farther down upon the cultivated land, and the river is broken by numerous falls, adorned with tall spruces, rearing their heads on the verges of precipices. On the third and uppermost terrace are extensive pastures, interspersed with chalets, but without any permanent dwellings. The solitary chapel of St. Guerin is raised on a knoll in the midst of the surrounding circus, and the paths over the Col of La Luce and the Cormet diverge from it across meadows and crumbling slate soaked with water. From the pass of the Grand Cormet there is a most splendid view of the snowy mountains of the Maurienne and Tarentaise, with extensive glaciers, and the fertile and broad valley watered by the Isere, and dotted with fine villages. The path descends to Aixme, where several Latin in- scriptions have been discovered which suggest its identity with Forum Claudii or Axima. The vale of Hauteluce is the broadest and least wild in the 190 Chaix's Notes on the Passage of Hannibal across the Alps. Beaufort country. It is fertile, and productive of wheat and hemp, while the forests merely crown the summits of the cultivated slopes. The cows are small but pretty, and the sheep yield a good fleece. E. of Hauteluce, the right slope of the valley is named Mon- tagne de la Ruelle, and the left slope Combe du Revers. The small river Dorinet is formed in the Combe a Dran, a barren gorge at the feet of the Aiguille Rousellette, the Fenetre, and the Plan de 1'Estace. It forms a waterfall where it leaves the Combe to enter the valley of Hauteluce, and it receives at the same place, through a more beautiful fall, the waters from the lake of La Girotta. I surveyed the lake. It is very picturesque, nearly surrounded by sloping pastures, with a few chalets, partly inclosed by dark, rocky cliffs, which give its clear waters dark-green, violet, and even black hues, according to the nature and depth of the bottom. I found the lake to be 1713 metres above sea ; the temperature of the waters 56° Fahr., that of the air being 50°, on the 15th Sept. 1853, at half-past 10 a.m. The only vestige of feudal times in these valleys is the ruins of a few towers on a hill N.W. of St. Maxime. According to the oldest memorials, the Beaufort country was a dependency of the archbishops of the Tarentaise. The feudal homage was due to them, when the princes of the House of Geneva acquired several tenures there ; and a deed dated July 31, 1220, recites that the Count of Geneva bound himself to render the yearly tribute of two large trout. There were, however, at the same time Lords of Beaufort, who, on the calends of April, 1271, sold most of their rights to Beatrice, the last Baroness of Faucigny. She erected the country into a Barony, which she united to her own, and both have since partaken of the same fortunes. In 1355 they were ceded to the Count of Savoy, Amadeus VI., by Charles, then Dauphin, and afterwards King Charles V. of France. When Henry IV. invaded Savoy, he came to St. Maxime de Beaufort with a body of troops on the 10th of October, and quartered in the castle. On the 11th he marched up to the pass of Cormet, to prevent Duke Charles Emmanuel of Savoy from entering the country from Piedmont over the Lesser St. Bernard ; but he returned on seeing that all the passes were impracticable, being already blocked up with snow, and he left Beaufort on the 12th. Agriculture affords but a limited resource to the inhabitants of these valleys. Saw-mills are numerous, especially at St. Maxime. The breeding of cattle and fine mules is the chief occupation. Cheese is also made on a large scale, and either sold to Piedmon- tese or bartered for rice and salt. The inhabitants are healthy, active, hard-working, and abste- mious. Their features bespeak intelligence. The women are not •wife Aft I'rtiiffuiifA. Sketch <>i'tli#- YAL1UBY OF B K.\ D FORT 9 &- Ihe iuljarnrt portions of Upper Savov. Sc al* 1 7.'.<>.i>"" ,;».,-. Wa* ,,: ' ' *W»fi/iv L'i> Kilometres 3 French Zeaanejr * * * *- - — * A" * * i A if A w,,/, jo* I»„hti'lf>t ,i'i <}i/ • »'..,ir/i, /,„ j;,/ (M ^ ltri „, . h^t.i.'u . lH.'>v Sawkins on the Volcanic Mountains of Hawaii, 191 destitute of beauty, and they are remarkable for an elegant head- dress. The preservation of their roads demands their constant exertion, owing to frequent floods ; but they are never in a satisfactory state. There is a tolerable inn kept by Henry Martin at St. Maxime ; those at Areche and Haiiteluce are very indifferent. The cot of Cyril Freyzon, at La Gite, can shelter the traveller bent on crossing the passes of La Perriere, La Sauce, and the Plan de l'Estace ; but Lavocat's chalet should, be avoided. XII. — On the Volcanic Mountains of Hawaii, Sandwich Islands. By J. G. Sawkins, Esq. Communicated by Sir Rod. I. Murchison. Read, June 25, 1855. There are four volcanic mountains in the island of Hawaii, viz., Kohala, Muna Kea, Hualalai, and Muna Loa. The decomposed state of the rock being greater on Kohala than on any other, and the total absence of black lava, lead me to believe that it is the oldest of the four. I also think that Muna Kea is next in age, and that Hualalai and Muna Loa are the most recent and of the same period. The height of these mountains above the sea is estimated as follows : — Kohala about 9,800 feet. Muna Kea „ 13,842 „ Hualalai „ 11,020 „ Mima Loa „ 13,651 „ I landed on the western side of the island at a place called Kawaihai, and crossed over to the eastern coast. This gave mc an excellent opportunity of seeing the decomposing effects of the moist atmosphere on the old lavas of the more northern and older volcanoes, which are no doubt co-existent with the forma- tion of the island. The most remarkable features on the N.E. portion of the island are the immense cliffs of compact lava, separated by ravines, varying in depth from 200 to 2000 feet, over the sides of which a series of cascades fall between Niulii and Hilo (or Byron's Bay). To the S. of that bay the lava assumes quite a different appear- ance ; the colour is black, and its structure like the slag of a fur- nace, having no definite form, though its cleavage is vertical. This black or recent lava extends itself into the ocean, forming a shelf, as though it had spread out horizontally from the coast.