AM~~~~~~~~~~~~~~~~~~~~~M Al~ ~~~~l Ali P Wal $EvXN 1|||_1111111111111111 11'11111 C' ' 1~~~~~~~~~~~Ha I ~ ' "ggaI~~~~tIIII~~~llllllllx~~~llll-ll l i l~~~~lli l'; '~~ rBz tjJt3t Kj g1 j, } jj jjll c.j 3................................ Ej ij j xrB jj.jlj ja W. 1............................., 4 4 Q ^ ^. c l r | ||| '. c..W|| |ll 1 11111111 ilillililil... 11111 Ilililli 1.lil l. | 1 i mil SE~f# —L f 0 [-<> / Jl >9 ' t3 ), S; i > f t 1 j~~~~~~~~~~~i Ii j~~~~~~~~~ - r.). ``, t //, ';2 \ X THE PRODUCTS AND STRUCTURE OF KILAUEA BY JOHN B. STONE BERNICE P. BISHOP MUSEUM BULLETIN 33 PUBLISHED BY THE MUSEUM HONOLULU, HAWAII 1926 CONTENTS PAGE, (Geographic features of the Kilauea regionl........-.............-................ 3 Topography-.......-................................. 3 Climate and vegetation.................................-.-. 6 Population................-.... --- —.......... --- —--------------------------------------------------------------------------—. Previous work........................................ ------------------ -------- ------------------ Acknowledgments..............................-........ 8 Lavas............9....................... 9 1 ntroduction.9.................................... --- —------------------------- ---- --—.................................... 9 Classification......................................9...................... g Pre-Kilauea series...............................-....... II Kilauea series................................................................................. I2 I'lows in the walls of Kilauea and the pit craters..-.............'...3..................... I3 The older surface flows.........-.............................. 5 The historic flows................-...6............................................. T6 Intrusive bodies in the Kilauea series........................................ I7 Ejected blocks.-.............................................................. 18 General petrography of Kilauea...................................................... 21 The ejected products............................................................... 23 Classification........................................ 23 Ashes in the pre-Kilauea series.............................-.-.-....... 23 Pahala ash...................... 24 Ashes of Kilauea crater............................-.............. 27 The Uwekahuna ash.....-.................-............... 27 O lder surface ashes.......................................................................................... 28 A\sh of 1924........................................................-.-........... 35 Structure...........6.......................... 36 K ilauea domne........................................ 36 Kilauea-,Mauna Loa contact.................-37.....................- 37 Cliffs of southern Puna and IKau-............................................................... 41 Puna ridge....................................................................... 42 K ilauea and the rift lines.................................... --- —------------------ ----------------- ------------ 43 Conclusions....................i....-s-............... 49 The geological history of Kilauea......................................................... 49 Comparison of Kilauea with other volcanic regions....................................... - 52. Bibliographyl.................. -........................... IILUSTRATIONS PAGE Plate.1. Eastern part of tlhe Uwekahuna instrusive body'; fossil footprints in the a sh o f 7....................................................... 59 Plate 11. Open fissure southwest of Kilauea....................................... --. 59 Ir(;URI TI. Alap of southeastern Hawaii..............-4...................... 4 2. Block diagram of the Puna ridge...................................... --- —-------------- ---------------------------- 3. MAap showing distribution of ashes around Kilauea.........................-.............. 29 4. MIap showing distribution of basal thread-lace scoriae.................................... 30 5. Structure sections.......................... -- - ----------------------- 38 6. Block diagram of coast south of Kilauea..-...................4..................4 7. Block diagram of Kilauea crater -....................................- 44 The Products and Structure of Kilauea BY JOHN B. STONE GEOGRAPHIC FEATURES OF THE KILAUEA REGION TOPOGRAPHY Kilauea crater is in the southeastern part of the island of Hawaii, 30 miles southwest of the harbor and city of Hilo. West-northwest of the crater is Mauna Loa (elevation 13,675 feet), an enormous gently-sloping dome 22 miles away; to the north is the cone-covered summit of Mauna Kea (elevation 13,784 feet), 32 miles away; to the southwest are the buttes of Mohokea, and still farther on is the ocean. Lava flows from Kilauea and its subordinate cones and fissures cover a four-sided area 52 miles long and i8 miles wide and comprising that part of the island of Hawaii lying southeast of a line extending approximately from Punaluu to the coast, about six miles southeast of Hilo. (See fig. i.) Within the area covered by lava flows from Kilauea, there has been practically no erosion. Rain sinks so rapidly through the fissured, porous, and cavernous lavas that the water table is very low. There are no permanent streams or springs even in the rainy part of the area, although brackish water can be found along the coast in a few cracks that reach down to sea level. Nevertheless, the unconsolidated ash deposits have been considerably eroded except where they are protected by vegetation. The torrential Kona rains cut trenches in the ash-covered slopes, and the strong trade winds are continually carrying the finer sand southwestward, where it forms widespread deposits. On days of unusually strong wind, dust clouds are carried out a mile or more over the ocean. In southern Kau below the great Hilina pali, an area a mile and a half long at its maximum about half a mile wide, is covered to a depth of several feet by angular bowlders and gravel which have been brought down by storm-water and landslides from the pali. Some of the fine sand from this bowlder plain has been blown to the southwest, where it forms sand dunes, many of which have the form of barchanes. Gravity is the chief transporting agent in many parts of the Kau desert and is rendered especially efficient by the fissured character of the lava flows. In some places the surface is littered with angular frag 4 Bernzice P. Bishop Museumt-Bulletilt 33 ments, many about the size of paving blocks, which make walking very tiresome. But the total topographic effect of this erosion is insignificant. All the major features of the topography and most of the minor features are those of lava eruptions and faulting. To an observer standing at the Uwekahuna triangulation station (elevation 4,090 feet) it is at once apparent that Kilauea is separate from Mauna " / / C^ ( ^i,1-6 ^ / ^ a. 'I 1-f i FIGURE i. Map of southeastern Hawaii. Lettered lines A-A, B-B, C-C, D-D, show the position of cross sections described in figure 5. Loa. Between Uwekahuna and the Mauna Loa slope is a broad, open valley, down which the pre-historic Keamoku flow turned after it reached the Kilauea boundary. (See fig. i.) A sudden change of slope at the intersection of the Mauna Loa and Kilauea domes all the way to the ocean is a conspicuous feature of the topographic maps. This topographic boundary is followed by the Kau-Volcano House road from a point about two miles northeast of Pahala to the Kapapala "halfway house." In many places along this road the nearly flat lava fields of Kilauea abut against a small pali' at the foot of the Mauna Loa slope. In places two or more 1 A number of Hawaiian words appear as proper names on the topogralphic maps of the United States Geological Survey. Common ones are: kipuka (an island-like area surrounded by one or more lava flows), puan (hill), and lua (pit or crater). The terms pahoehoe and aa distinguish two types of lava that are in general use in geological writing and (do not need definition lhere. Stone-Kilaulca 5 palis form a terraced slope, in others the palis have been almost completely obliterated by flows from Mauna Loa. North of Kilauea crater the bounding escarpment dies away near the crest of the Kilauea dome, and from that point to the seacoast southeast of Hilo there is no marked topographic break between Kilauea and Mauna Loa. Kilauea crater itself is a depressed area in the summit of a dome, which slopes away in all directions. It is shut in on all sides except the south by perpendicular cliffs which reach a maximum height of 450 feet on the northwest side. In places, notably southwest of Uwekahuna and at the Volcano House, are terraces formed by fault blocks. The south rim of the crater is a low cliff partly concealed by ash deposits. At two points the south wall of the depression is very low, and in I92I short streams of lava from Kilauea overflowed into the Kau desert. The crater floor is itself a gently FIGU;R 2. Block diagram of the Puna riidge showing pit craters, cones, fissures, and fault cliffs. sloping lava dome having at its summit the yawning pit of Halemnaulmau, 3,400 by 3,000 feet across and 1,350 feet deep (in 1925). The only irregularities of the floor are minor domes and a few small spatter cones on lava flows. Extending southeast from Kilauea and gradually curving around to the northeast is a broad ridge or nose reaching all the way to Cape Kumukahi, the easternmost point of Hawaii. (See fig. 2.) Following the summit of this ridge quite closely is a line of pit craters, lava cones, and fissures. The pit craters are straight-sided depressions usually with a roughly circular or elliptical outline. They are all subsidence craters formed by collapse in a flat terrane and are not summit craters of cones. Their ulpper walls are cliffs composed of horizontal lava flows; the bottoms of most of them are talus funnels, but a few have level floors formed by the cooling of lava lakes. The line beginning with Keanakakoi includes twelve of these pits /C ~ `~- ---— 1_4~/ --- —-,~tn 6 B1crn icc P. Bishop Mluscu-zBulletinl 33 ranging in dimensions from the Devil's Throat, 35 feet in diameter and 250 feet deep, to Makaopuhi, a mile long and nearly I,ooo feet deep. Kilauea Iki, another large crater, lies east of Kilauea, not in the chain of pits. Most of the lava cones and cinder cones lie east of the pit craters. The largest is the 400-foot cinder cone at Kapoho. Another large one is the flat dome of Kane Nui o Hamo, a cone 300 feet high with a summit crater as deep as the cone is high. Following the line of craters and cones through Puna are many series of parallel cracks, gaping fissures, some as much as 50 feet across. From Kilauea southwest to the sea the country slopes quite regularly. Striking parallel with the slope are many open fissures, along some of which there has been enough vertical displacement to form small cliffs. One such cliff, Puu Nahala, is somewhat over oo00 feet high. Several lava and cindercones are conspicuous features along this rift line. The Kamakaia Hills are especially large and symmetrical, the largest being 150 feet high. In detail the surface is rough because of its covering of fresh flows, either hummocky, irregular pahoehoe or jagged, bristling aa. For about six miles south of Kilauea the country has a regular southerly slope, interrupted only by a few small inward-facing fault cliffs. In the remaining three miles or so to the sea the land falls off about 2,000 feet in a series of giant terraces bordered by cliffs as much as I,500 feet high. The coast-line itself is nearly everywhere a cliff, usually 20 or 30 feet high, blut at Puu Kapukapu I,oo0 feet high. Most of the smaller cliffs are covered l)y cascading lava flows. Aa flows are especially numerous coming down over the cliffs in Puna. Wave elucsioll is now cutting back the coast, but its total effect is slight. Deep water comes to the shore in most places. CLIMATIE AND VEGETATION lThe climate of the K-ilauea-lKau-Puna region varies remarkably. The prevailing northeast trade winds bring abundant rain to the portion of the area north of the Puna ridge. At Hilo the average annual rainfall is 139 inches, but it decreases at altitudes above 3,200 feet and is only 88 inches at the Volcano House. Humidity is high, at the Volcano House often 100 per cent. In striking contrast is the leewar( portion of the area, south of the line. In the Kaau desert southwest of Kilauea the rainfall is so small that vegetation is nearly absent and dust storms may be seen when it is raining at the Volcano House three miles away. In general the parts of the area that slope to the south and southwest are poorly watered. What rain they receive is mostly brought by the southwest or "Kona" winds which blow occasionally, especially during the winter. Sto tc-Kila tca 17 The temperature of Hawaii is equalle and in spite of the sul-tropical latitude is never extremely high. At Hilo the mean annual temperature is 720 with a range fronm 65~ to 79.5~ (mean minimum and maximum). Temperature decreases with increasing altitude so that the air at the Volcano House is usually invigorating and the nights are uniformly cool. The mean annual temperature is 6 1~ ranging from 54~ to 68~.,ight frosts occur rarely. The climatic variation is strikingly evident in the vegetation. Th'e whole windward slope is covered by a dense forest, in many l)laces impenetralle except along trails. At lower altitudes the forest is made up of strictly tropical trees, among which are the kukui, the hala or screw-pine, the b)anana, the breadfruit, the mango, and the guava. Along the coast are coconut palms, in places in large groves. At higher altitudes this trolical forest is gradually replaced by tree-ferns and the lehua, which form nearly the whole forest above 2,000 feet. A few sandalwood trees and a few koa grow near the volcano, but the koa thrives better at higher altitudes. Climbing plants like the ieie and the uluhi (staghorn fern) form entanglements in many places. On the lee side of the area the forest thins rapidly. Along the south coast as far west as Kamoamoa are scattered groves of coconut palms and hala trees and thickets of guava bushes, but the coast west of lKamoamoa and nearly all the district of lKau are thinly wooded. Large areas are covered by barren lava flows with only a few scattered lehua trees. Soil is limited to patches of volcanic ash, but in all excel)t the most arid spots grass grows wherever it can take root. POPULA'0TION Eastern Puna has a number of inhabitants especially around the sugar pllantations near Kapoho and ZPahoa. Cattle are pasturedl on some of the lands unfit for cultivation. Only a few peol)le live in the rest of the area. Near the Volcano House are the military recreation camp and a number of summer cottages, but the triangular area between Punaluu, IKilauea crater, and Kalalana is entirely uninhabited and, except for cattle on the flats south of Pahala, is abandoned to the wild goats and dontkeys. PRIEVIOUS W\ORK Since 1823, when Kilauea was visited by a party of missionaries, there has been some sort of record of the activities of the volcano and since the establishment of the Hawaiian Volcano Observatory in 1912 the rec r(l has been detailed and complete. The work of the ()Observatory, however, has 8 Bernlicc P. Bishop Aluseuml-Bulletin. 33 been largely restricted to recording the present-day changes of the volcano; and most geologists who have visited Kilauea during the century since I823 have been forced b)y the shortness of the time at their disposal and by the lack of a large-scale topographic map to confine their observations to the immediate neighborhood of the crater. Consequently no detailed geological study of the whole area has yet appeared, although much has been written alout the volcanoes of Hawaii, and about Kilauea in particular. The most important writers treating of the general geology and petrology of Kilauea are: Brigham (4),2 Cross (8), Daly (Io), Dana (ii), Dutton (12), I-litchcock (17), Jaggar (20), and Washington (38). Purely volcanological papers are omitted. T'he first systematic geological work on Hawaii was begun in 1920 by L. F. Noble (28)" and W. 0. Clark in studying the water resources around Pahala, IKau district. Their work showed that the lavas in the Pahala district, which is on the flank of MIauna Loa, can be divided into three series. The first two series are separated by a profound unconformity. This successful demonstration of distinct epochs in the life of Mauna Loa offered the first definite hope for the solution of the geology of Hawaii. ACKNOWLEIDGMENTS During the summer and fall of 1925 I spent five months on the island of Hawaii, mostly in the area discussed in this paper. My study was made possible by a Bishop Museum Fellowship awarded by Yale University, and this p)aer lhas already been presented as a dissertation in candidacy for the degree of Doctor of Philosophy at Yale. The prol)lem was outlined by Professor Herlert E. Gregory. While in the field I received assistance from a great many friends, among them Mr. and Mrs. \V. F. Stephens, Mr. B. M. Sumner, and Mr. A. M. Brown, Jr., whose courtesy and that of many others it is a real pleasure to acknowledge. I am especially indebted to Dr. T. A. Jaggar, Jr., and Mr. R. IH. Finch of the Hawaiian Volcano Observatory, who did everything possible to help me and from whose suggestions I have profited greatly. Mr. W. O. Clark guidled me on several field trips to Kau and discussed some of the prol)lems with me. In preparing this lpal)er I have had the helpful advice of Professor Adolph IKnopf of Yale University. A number of thin sections were made for me 1)y the United States Geological Survey. Dr. C. K. Wentworth kindly loaned me his apparatus for constructing block diagrams. 2 Numbers iln parentleses refer to Bibliography on page 58. 3 In 1924, II. T. Stea-ns studied tlhe geology of the Kat district. Ilis report probably w-ill include the results of work by Noble, Clark, and otlers. StoncJ-Kilatica. 9 THE I AVAS I NTROI)UCTION Hawaii, like the other islands of the group, is entirely volcanic. \Vith the exception of the trachytes of Puuwaawaa and Puu Anahulu (7) and the possible oligoclasite of Waimea (38, p. 477), the lavas are all basaltic, for the most part ordinary feldspar blasalts and augite andesites but with some more basic varieties. Fragmental layers form a very small proportion of the island, and dikes or other intrusive bodies are few and small. There are five recognized centers of eruption, of which two, Kohala and Alauna Kea, have been inactive for at least several hundred years. Kohala seems to be much the older. Hualalai has been inactive since I8oi, but Mauna Ioa and Kilauea have been active during most of their recorded history, although both were quite dormant in 1925. The windward slopes of Kohala are dissected by deep gulches, and so to a less extent are those of Mauna Kea, but no study has been made of the rocks of the lowest flows. The other volcanoes are practically untouched by erosion so that the only sections exposed are in fault cliffs. It is generally assumed in discussions of Hawaii that the island has been built by the five recognized volcanoes, but this is by no means certain. An old land mass underlies Kilauea and at least part of Mauna Ioa. That this land mass was built by the present Mauna Ioa is quite possible but not Iroven. The evidence of extensive faulting along the southern coast and probably also along the northern side of Kohala suggest that there may have been older volcanic centers of which nothing is known, now silnk beneath the ocean or buried by the later volcanoes. CLASSIFICATION The classification of basic lavas has been treated with especial reference to Hawaii by Washington (38, pp. 465-474), whose conclusions are sunmmarized here. Lavas comlposed essentially of plagioclase and l)yroxene, with or without olivine, in the proportions in which they occur in Hawaiian rocks, may be divided into three great classes; andesite, basalt, and more basic rocks, depending on the relative proportions of salic and femic minerals. The arbitrary limits set for the groups 1y \Vashington are: andesite, salic minerals (almost entirely plagioclase) 87.5 to 62.5 per cent; basalt, salic minerals 62.5 to 37-5 per cent; more basic rocks, less than 37.5 per cent. The groups andesite and basalt are again divided on the basis of the composition of the plagioclase. When the plagioclase of a basalt is not speci 10 Bclrnicc P. Bishop:Tlulsclim-B lletinl 33 fied, it is assumed to be labradorite. Finally the presence or alsence of olivine makes possible another subdivision of these two groups; but the term "basalt" is not limited to olivine-bearing lavas. The more basic rocks are divided into two groups, one in which the percentage of salic minerals is between 37.5 and 12.5, and an ultrabasic group with less than 12.5 per cent of salic minerals. The group containing 37.5 to I2.5 per cent of salic minerals is represented in Hawaii only by a class of lavas rich in olivine; these lavas are called "picrite basalts." The ultrabasic group is not representel. It should be noted that in Washington's scheme the ratio of salic to femic minerals, as well as the composition of the plagioclase, is that calculated in the norm. The arlitrary limits set by Washington for the groups, such as andesite and basalt, may not have general validity, but they probably coincide closely with ordinary usage. For the rather uniform rocks of Kilauea the fundamental division into the three varieties, labradorite basalt, olivine-labradorite basalt, and plicrite basalt, has been well brought out by Washington's analyses and could not be brought out by any other method in the common glassy lavas. In the present paper I have classified lavas by correlation with analyses given by Washington. The method presents no difficulty when applied to the Kilauean lavas because the distinction between basalt and olivine basalt depends on a modal characteristic, the presence or absence of olivine; and the only rocks more basic than basalts are picrite basalts, which can also be recognized by their high content of modal olivine. Accordingly rocks with less than 2 per cent of olivine are classed as basalts, those with more than 2 per cent and less than I5 per cent as olivine basalts, and those with more than 15 per cent as picrite basalts. The boundary between olivine basalt and picrite basalt is somewhat uncertain as the only olivine basalt analyzed by Washington (38, pp. 342-344) has 7.47 per cent of olivine in the norm and about I2 per cent in the mode. The other analysis given in the same place as of an olivine basalt is that of a specimen collected b)y Cross (8, p. 42) and said by him to be poor in olivine. The coml)ositions of the plagioclases given in the present paper are those in the mode. Comparison of the mineral coml)ositions given in this paper and the norms calculated by \Vashington shows that the percentage of feldspar is much lower in the mode than in the norm; but the data are not sufficient to correlate the modal and normative compositions. It is 1)lainly desiral)le to letermine the actual mineral composition of rocks analyzed chemically, so that they can be used as standards of reference. Unfortunately micrometric measurements of basalts are all too rare in geological literature; the only previous ones for Hawaii are those given by Daly (9, pp. 292, 296, Stonet-Kilauica II 302). On the chemical and mineral composition of basalts of the Arctic region Holmes (I9) has reported. Kilauea is built of superposed lava flows with a few interbedded cdeposits of fragmental material. The flows are divided into two series; an older series, here called the pre-Kilauea series, consisting of horizontal flows with a conspicuous ash bed (the Pahala ash) at its tol; anid the younger or Kilauea series lying unformably on the pre-Kilauea series. T1-lE PRE-KIILAUIE\ SERI'S The pre-Kilauea series is exposed in some of the high cliffs in southern Puna and Kau. A section of nearly 1,500 feet is exposed in Hilina pali and a section of I,000 feet in the seaward face of Puu Kapukapu, but unfortunately most parts of these cliffs are inaccessible. The series consists of horizontal flows and ash beds cut by a few small dikes. The ash beds, which are thin and discontinuous, make up perhaps five per cent of the section. Aa flows with clinkery tops and a maximum thickness of about 20 feet predominate. The massive portions of the aa flows are light or medium gray, aphanitic rocks having a few small distorted vesicles. Some are aphyric except for very sparse rounded microphenocrysts of olivine. Another common type has feldspar phenocrysts as much as two millimeters long. A thin section of this type of rock showed a few microphenocrysts of a brownish augite in addition to laths or groups of laths of a sodic bytownite (Ab.,,An7,). The ground mass is very fine grained with intersertal texture and a very little glass. The ash at the top of the pre-Kilauea series is correlated with the Pahala ash of western Kau. (See p. 23.) The underlying flows and ash beds therefore correspond in stratigraphic position with the Pahala and pre-Pahala series as defined by Nolle and Clark. In the cliffs of eastern Kau and western Puna, however, there is no break in the older lavas, nor are any ashes interbedded with the later flows, but the uppermost bed of ash appears to be faulted. The Pahala series in eastern Kau apparently lies conformably on the pre-Pahala series. The following description of the type pre-Pahala and Pahala series is based on Noble's manuscript report, as quoted by Washington (38, pp. 119-121):.... the pre-Pahala series consists chiefly of massive, rather uniformly bedded flows, most of which range in thickness from 6 to 25 feet. The series includes a few beds of stratified yellow ash, none over 15 feet thick and most much thilnnr. 'The prevailing color of the lavas is light bluisl or piikish gray contrastinlg with the very (lark recent and historic lavas. Mluch the greate r number blelong to oile gclneral ty pe, consisting mainly of plagioclase, augite. and subordinate magnetite, with olivile rather sparingly present. 12 Bern ice P. Bishop Aluselum —B lletil 33 Lavas of the Pahala series are in general medium to rather dark bluish-gray rocks, lighter in color than most of the historic and recent flows, but much darker than most of the pre-Pahala lavas, and lacking the pinkish tints that many of the pre-Pahala rocks show. The dominant type is plagioclase-augite basalt moderately poor to moderately rich in olivine. The Pahala series is characterized by beds of stratified yellow ash, one at least 75 feet thick in places, another at least 50 feet. The Pahala series is exposed at a numler of places along the northwest edge of the Kilauea area. The pre-Pahala series is exposed in the walls of Mohokea and Wood Valley. The description by Noble of the prePahala series agrees well with the rocks in the lower part of the cliffs of eastern ]Kau. In the upper third of the cliffs ash beds are more important, and most of the flows examined are pahoehoe lavas much like the dark lavas of the lKilauea series. This agrees with the description of the Pahala series. The source of the pre-Pahala and Pahala series, and therefore of the corresponding pre-Kilauea series, was a volcano in the general vicinity of Mauna Ioa. This origin is proved by the dips in the Pahala region and by the olservations of Clark (5) that the older series of lavas are found on the slopes of Mauna Loa up to an elevation of at least 7,500 feet. TIHE KILAUIEA SERIES The IKilauea series includes all lavas and ash beds of the Kilauea area which are younger than the Pahala ash, with the exception of some postPahala flows from Mauna Loa, which interfinger with the Kilaueau flows along the contact of the two domes. The Kilauea series covers the surface of the entire area except the cliffs and kipukas where the pre-Kilauea series is exposed, and is well displayed in a number of places (fig. i). A section 450 feet thick appears in the walls of Kilauea crater and other sections are exposed in the pit-craters of Puna, the thickest being 700 feet in the crater of MAakaopuhi. The Kilauea series probably has a maximum thickness of at least I,ooo feet. In the southern part of the area, however, there is little evidence of the thickness. Sections as nuch as Ioo feet thick in the face of Puu Nahaha and in the (Great Crack do not expose the base of the IKilauea series. In some places along the tops of the big cliffs the Kilauea flows overlying the Pahala ash have a total thickness of less than 00o feet. W\ith the exception of the Uwekahuna ash and the surface ash around the crater, the lKilauea series consists entirely of lavas, pahoehoe flows being much more numerous than aa. Complete mapping of the individual flows is iImpracticable because of the ash and forest cover, which hides a large part of the region. It is extremely difficult to separate pahoehoe flows even in the desert country when once the younger ones have lost their fresh appearance. The topographic maps of the United States Geological Survey show Stonc -K illauca 3 all the historic flows and also most of the areas covered )y aa. although individual aa flows as a rule are not outlined. The flows mapped vary greatly in size, ranging from the flow of I868 with an area of about 31 acres to the great floods of 1823, I920, and the IKamooalii flow with areas of 5, 5.2, and about IO square miles, respectively. The older flows all)arently had a similar range of size. The flows of the Kilauea series can be divided into a numl)er of groups depending on their relative ages and distribution. FLOWS IN TJIE WVALLS OF KILAUEA AND TIIlE PIT CRATIRS The oldest exposed lavas of the Kilauea series make up the walls of the craters. They are almost exclusively pahoehoe flows with a maximum thickness of about 30 feet, but aa flows occur in several of the pit craters. There are a few dikes and irregular intrusive bodies. A thin fragmental bed, the Uwekahuna ash, is exposed for a distance of 300ooo feet along the northwest edge of the floor of Kilauea, but no other interbedded ash deposits are exposed in Kilauea or in the pit craters. The lavas of the walls of Kilauea are in general medium to dark gray, vesicular, aphanitic rocks. The lower parts of the thicker flows are nonvesicular and in general lighter in color than the vesicular lavas. Nearly all the lavas have recognizable phenocrysts of feldspar and olivine, but olivine is not abundant. In thin sections of these lavas olivine appears only as small, sparse phenocrysts. Plagioclase feldspar having the composition of a calcic labradorite (average Ab:.3An(.,) forms microphenocrysts in nearly all specimens, and brownish augite in several. The groundmasses consist of calcic labradorite, augite, magnetite, ilmenite, and glass. The texture of the nearly holocrystalline specimens from the centers of the thicker flows is intergranular or intersertal. Augite forms elongated, ragged grains between the diverging laths of feldspar; iron ore occurs either in shapeless grains accompanying the augite and feldspar or, more commonly, as crystal growths in the brownish primary glass, to which it gives a black, opaque appearance. Very glassy specimens have a black groundniass containing faint, feathery microlites of feldspar. A phaneric specimen from the cliff at Uwekahuna was measured by the Rosiwal method. The specimen is a nonvesicular rock with an irregular variation in texture; parts are medium-gray and aphanitic except for some specks of feldspar; other parts are grayish-brown, porous, and crystalline. Sparse phenocrysts of olivine have a maximum diameter of 2 millimeters. Under the microscope the rock was found to consist of augite, calcic labra I4 Bernice P. Bishop liuseum-1Bulletin 33 dorite (about Ab:;,An05), iron ores (mostly magnetite), and a few small patches of glass. The result of the measurement expressed in weight percentages is: Olivine................................... I.i A ugite...................................... 54.2 Labradorite...................... 31.3 Iron ores.............................. 2.6 Glass........-................. o.8 I00.0 Computed sp. gr.................. 3.07 Measured sp. gr............... 2.9 (The rock is quite porous.) Specimens collected by me from the walls of Kilauea are all of labradorite basalts very poor in olivine. Washington found labradorite basalt the most common type in the crater walls but also found a few flows of picrite basalts. Flows in the walls of the pit craters are similar to those at Kilauea, but one flow in the crater of Makaopuhi deserves special mention. Makaopuhi is a double crater, whose older eastern pit had been partly filled by lava before the deeper western pit fell in and exposed a section through the filling. At the top of this section are four pahoehoe flows with a total thickness of about 20 feet; below the flows is a massive body of lava at least 150 feet thick resting upon the reddened slope of the eastern pit. This massive body is a thick, plano-convex lens, which filled the lower part of the crater. It has good prismatic jointing, which is vertical in the upper part of the fill but in the lower part is perpendicular to the convex base. Iight green blrochantite coats some of the joint planes. The rock of this fill is light gray and phaneric with a few small phenocrysts of olivine. It consists of augite, labradorite (Ab40An,6), iron ores, and olivine with a little brown glass containing fine needles of apatite. The texture is intergranular, but locally there are diabasic clots. The mineral composition in weight percentages is: Olivint e...................-....................... ---. 1.T A ugite........................ ----. --- —---......... 48.2 Labradorite........................................................ 35.0 Iron ores...................................................... 13.2 G l a ss......................................................................................... 2.5 100.0 Computed sp. gr.......-......-..-..........7..... ----.......07 M easured sp. gr.....................................-.......................... 3.03 Stone —Kiltauca' I5 TiHl OLDE)R SURA.\Ct: FLOWS The only flows of the Kilauea series exposed at most places away from the craters are those at the surface. Some of these correspond in age with the topmost flows of the crater walls, but others are the products of later fissure-eruptions. The only means of dating most of them is their relation to the older surface ashes, but the ashes do not extend far from Kilauea. Only the historic flow can be definitely set apart. In general appearance the pre-historic surface flows are the same as those of the crater walls. Dark gray, vesicular, aphanitic lavas are most common, but in thicker flows the lava is quite crystalline about three feet below the surface. Many of the lavas have feldspar phenocrysts as much as 2 millimeters long, and a few flows are moderately rich in olivine. Thin sections of these lavas all contain some olivine, but in most sections the olivine grains are small and scattered. The flow of Puu-kole and the big aa flow at the foot of the Hilina pali, however, are fairly rich in olivine. Several specimens have the elongated, rod-like olivine phenocrysts described by E. S. Dana (II, pp 324, 326). The plagioclase feldspar is a labradorite at least as calcic as Ah,,4AnO,, and in a few sections the phenocrysts are sodic bytownite. Both feldspar and augite occur in two generations. The groundmasses of most of these lavas consist of feldspar and augite in a black, glassy matrix. The feldspar is in the usual lath-shaped crystals with irregular orientation, and the augite is in anhedral grains accompanying the feldspar and showing a preference for growing along or around it. Under a high powered objective the opaque matrix or base is resolved into a clear brown glass full of beautiful crystalline growths of magnetite and possibly ilmenite. The magnetite is in sharp octahedra arranged in branching lines. All the octahedra in a line have one crystallographic axis in common, and the various lines are parallel to the isometric axes. IExactly similar crystallites of magnetite in basalts from Nevada were described and figured by Zirkel (42). In more nearly holocrystalline specimens the ores are not confined to the glassy mesostasis but are distriluted through the rock, partly in parallel rods. A compact, nearly holocrystalline specimen from near the base of a 20-foot flow was measured by the Rosiwal method. The section consists of a few olivine phenocrysts in a diabasic groundmass of augite, lalradorite, iron ores, and some brown glass. The proportions are expressed in weight percentages. Beniice P. Bishop M11usctum-Bulletin 33 01 ivine...-..-... ----.. --- —-..... —........ -—......................... 6.7 Augite... -................................... 44.1 Labradorite...... --- —-.... --- —.. ---.. ----. --- —---—...-...-..-.......................... --- —----------------—............ 28. Iron ores................................................................ 15. (G lass................................................................... 6. 100.0 Computed sp. gr........................................ 3.25 Measured sp. gr...................................................... 2.91 THEI HISTORIC FLOWS The historic flows of the Kilauea series include flows from fissures along the northeast and southwest rift lines and flows from Halemaumau. Flows were erupted along the rift lines in 1823, I840, I868, 1920, 1922, and 1923; in Kilauea Iki in I832 and 1868; and in Keanakakoi in I877. The flow of 1823 (the Keaiwa flow) came from the southern half of a fissure eleven miles long southwest of Kilauea (35, 37). The lava is a dark gray, vesicular basalt with recognizable phenocrysts of feldspar and olivine. Augite microphenocrysts are also seen under the microscope. In 1840 lava broke out at several places in the neighlorhood of Makaopluhi, but the greatest outburst was the Nanawale flow in eastern Puna. The lava of this flow has been described by Cross and Washington and analyzed by Steiger. It differs from most other flows of Kilauea in being highly chrysophyric. The groundmass consists in order of abundance of augite, highly calcic labradorite, magnetite, and glass (Cross). The lava of i868 reached the surface southwest of Kilauea in four little patches, of which the largest is only I,000 feet long and the others are much smaller. It is of the common type of Kilauean lava consisting of phenocrysts of olivine, lalradorite, and augite in an opaque black base. The erupltion of I920 is described in great detail in the Monthly Bulletins of the Hawaiian Volcano Observatory for that year (21, 22). A lava dome, Mauna Iki, about 125 feet high was built in the Kau desert southwest of Kilauea, lava flows extending for more than six miles off to the south. Both aa and pahoehoe lava were formed. Specimens of each variety have been analyzed by Washington; they are of the common type of lalradorite basalt poor in olivine. In I922 a small flow broke out of a fissure in the wall of the western pit of Makaopuhi and covered the bottom of the crater with a jagged mass of aa and rough pahoehoe. Small flows also apl)eared at Napau crater. Again in 1923 lava appeared near Mlakaopuhi. All these lavas are of the comm1on type, as are also the historic flows of Kilauea Iki and Keanakakoi. The floor of Kilauea crater at the present time is almost entirely cov Stonc-Kilaituca 17 ered by the flows of i919 and 1921, which were formed l)y overflows of Halemaumau. Older lava, probably of 1894, is exposed in places and has been analyzed by Ferguson.4 Many other historic flows are exposed in the walls of Halemaumau. The flow of 1919 has a few comparatively large phenocrysts of olivine and smaller ones of augite and lalradorite (A\l4,AnI, or more calcic). IN'rIzUSIV\ BoIIKS IN TIII IKILA\UI:A SKRIKS A few small dikes and irregular intrustive bodies occur in the walls of the craters and are especially prominent in Halemaumau. The rock of the dikes is in general dark gray to black, aphanitic basalt, and an analysis by Washington of a specimen from a dike in the north wall of Kilauea showed that the chemical composition of the dike rock is the same as that of many flows. The only analysis of one of the larger intrustives can also be matched among the flows. A conspicuous body of massive rock in the wall of Kilauea northeast of the Uwekahuna fault terraces was examined by Daly (9, pP. 291-292; Io, pp. 1 5- I6) and considered a laccolith, but Daly's conclusion has not been generally accepted. Most geologists have followed Powers (32, p. 33) in believing that the massive rock is the filling of a former lava tube. The body can be more readily examined now that the flow of 1919 has raised the floor of the crater, and small avalanches have cut into the wall. The "laccolith" consists of two plano-convex lenses placed end to end with their flat sides down and at nearly the same level, having a total length of about 750 feet and a maximum thickness of 40 feet. (See PI. I, A.) The mass lies with very slight disconformities on Io to 15 feet of peculiar red, scoriaceous lava flows averaging 2 inches to 6 inches thick but including a few flows I foot or more thick. Above the lenses of massive rock are more of these thin flows and a few heavy flows, which arch up over the north end of the northern lens. The remaining height of the 450-foot cliff consists of massive, horizontal flows. Landslides have cut into the south end of the lens about 40 feet and show the base dipping 5~ into the wall. Other landslides have broken the surface connection between the two lenses. The flows both above and below the lenses are somewhat brecciated; small apophyses cut the overlying rock; and there are irregular patches of what appears to be intrusive rock in the underlying flows betweenl the two lenses. A two-foot dike of material like that of the main body cuts the underlying flows and the southern end of the northern lens. The top of the massive rock is non-vesicular, and the base has a chilled nmargin and 4.ll the reliable analyses of Kilauean lavas are listed by Washington (38, Ip. 342-352). Bernlice P. Bishop Mluscunz-Bullctin 33 upright tuluar vesicles. The l)ack wall of the northern lens where exposed is nearly vertical. The intrusive relations at the base of the "laccolith" are opposed to the lava-tube theory, and it seems probable that the body may be an irregular intrusive injected under thin cover near a lava lake. The intrusive bodies in Halemaumnau are examples of masses of this origin. The thin, scoriaceous flows associated with the massive lenses were prolably formed by thin, hot overflows from a lake. The rock of the Uwekahuna intrusive body is a gray, crystalline, gabl)roid rock, very rich in olivine. The mineral composition calculated from the chemical composition and a micrometric measurement is given by Daly (9, p. 293) as: Olivine................. --- — -------—....................... 40.0 A ugite..........-. ---..-.............-.......................................................... 31i.o Labradorite (AbhAnmI)........................................ 27.0 Magnetite and Ilmenite.....-..................... -.7 Apatite......................................................................... --- —---------------------------------------- ----------—................ 0.3 100.0 Sp. gr........................................................................................... 3.oo EJT'CTIrD BLOCKS Thle floor of the crater around Halemaumau is thickly strewn with rock fragments of all sizes thrown from the pit during the explosive eruption of May, 1924. On the surface of the "sand-spit," bowlders of an earlier eruption-that of 179o-are mingled with the more recent ones, and lowlders are found in all the ash deposits near Kilauea. Petrographic descriptions of the ejected blocks are given here for comparison with the lava flows; the ash deposits as a whole are described( as pages 23-27. The ejected blocks are entirely of basaltic composition. A great many are of rock types like the recent lavas; others are red, vesicular lava, which is l)rolbaly)l old talus or crag material from the former lava lake; but there are many blockls quite unlike the surface flows. Specimens from the ejected blocks of the older eruptions have been described by Dana, Daly, Cross, and \\ashington. The thin sections here descril)ed are from the blocks of the etrultion of 1924. A coullmon and consplicuous variety among the ejected blocks is a light gray, vesicular rock with large p)henocrysts of glassy, yellowish green olivine in an aphanitic groundmass. As commonly true of HIawaiian lavas, the vesicles are miarolitic cavities containing fine acicular crystals and thin tables of feldslar. Olivine makes up 23 per cent by volume of the rock. The texture of the groundmass is unusual and consists of small laths of Sftonc -Kilauca I9 feldspar, partly in sheaf-like groups, surrrounded by grains of augite, which is extraordinarily abundant. Iron ore is distributed through phenocrysts and groundmass in irregular grains. A somewhat similar texture was found by ~Washington (38, p. 344) il one of the older blocks. Another olivine-rich specimen was selected for measurement by the Rosiwal method. It has large phenocrysts of olivine in an intergranular groundmass of augite, labradorite (Ab;,,An,,;), iron ores, and a little glass. The olivine is full of tiny rods or spindles of a faint purple color, arranged at right angles to the elongation of the olivine. The mineral percentages by weight are: O livine -.......-... --- -—.. —...... --- —. — ---------... --- ----------------. ------ ------------ - 31.8 A ugite..............................................................................38.9 L abradorite.................................................................... 23.4 Jron ore.............-.......................... 3.8 Glass...................................... 2.1 100.0 This composition corresponds with a specific gravity of 3.2 compared with the measured specific gravity of 3.04, but the rock is vesicular and porous. Two specimens of rock rich in olivine also contained hypersthene. In one specimen the hypersthene forms rounded phenocrysts. The other specimen, which is of a rock with perfect platy jointing, probably from the large intrusive body in the wall of Halemaumau, contains large poikilitic anhedra of hypersthene inclosing feldspar laths and a little olivine, measurement of which gave the following mineral composition in weight percentages: Olivine.......9............................... ---. 19.0 Hypersthene................................... - ---------- - 5.9 Augite............................. --- —-------- - 45. Labradorite.............................................22. 1 Iron ore......................................................... 6.6 (Glass...............................3 100.0 Computed sp. gr....................-.................... —.. 3.2 M easured sp. gr..........................-... - --------—. ------ 3.09 Another rocl type differing from the ordinary flows of the Kilauea series is light-colored, dense, aphanitic lalradorite basalt with a very little olivine. A specimen of a light brownish gray rock has a few small olivine phenocrysts and a few specks of feldspar, which the microscope shows to be groups of labradorite laths (Ab,,oAn,,,) in an intersertal groundmass of augite, labradorite, and iron ores. 20 0Bcrnicc P. Bishop Mlluseuc —l-Bulletinl 33 The vesicles in many of the ejected blocks are true miaroles lined by projecting crystals. Feldspar, augite, ilmenite, and magnetite can be recognized with a hand lens. In several specimens of the ejected blocks and also in specimens from the Uwekahuna "laccolith" many of the vesicles contain little, round, white crystals about 0.2 millimeters across, which were noticed by Dana ( I, Ip. 327-328) and Washington (38, p. 340) and considered a zeolite. Upon examination with a binocular microscope, nearly all these were found to be complexly twinned, but some have faces of either octahedra or dodecahedra, and many have hexagonal outlines. The mineral was examined by the immersion method and found to have very low birefringence, an irregular mottled appearance between crossed nicols, and an index of refraction very nearly 1.485. These optical properties indicate either analcite or cristobalite. The mineral was infusible before the blowpipe, gave no sodium flame, and was not fusible with hydrocholoric acid. Mr. J. F. Schairer kindly tested a small amount of the mineral weighing alout 8 milligrams and found 75 per cent of silica. The impurity is probably due to small fragments of feldspar and ilmenite attached to the crystals. The mineral, therefore, is quite surely cristobalite. Cristobalite was definitely identified by Cross (8, 1). I) in an olivineplagioclase basalt from Olokele Canyon, Kauai. It occurs in vesicles as minute, dull-white rhomlbic dodecahedra about 0.3 to 0.5 millimeter in diameter, and has the same optical properties as in the rocks of Kilauea. A few of the ejected blocks are slightly altered. The olivine phenocrysts in one sl)ecimen contain a red substance, in places in regular, branching patterns, and in other places so abundant that the olivine is opaque. The augite is somewhat discolored, but the feldspar is perfectly fresh. In the same specimen the vesicles are lined by a pale blue coating, which under the microscope is a cryptocrystalline aggregate with moderately high birefringence. The same substance fills some of the interstitial spaces in the groundmass and has possil)ly replaced primary glass. Some of the more coarse grained blocks crumlle easily, probably as a result of having been reheated. Many of the bowlders ejected in 1924 were incandescent. Ejected blocks are also found around the pit crater of Alealea (Alae) and along the source fissure of the flow of I823 in Kau. Two specimens from Kau were studied in thin sections. One is a light gray, nonvesicular phaneric rock, a typical basalt with a few small olivine phenocrysts. Some of the iron ore is in rods or pllates having a perfect parallel arrangement. The other specimen is an aphanitic, vesicular rock of an unusual light brown color. It has a few rounded microphenocrysts of olivine in a very fine grained, hypocrystalline groundmass. Stonc-Kilaitca2 21 Mlost of the blocks are plainly from lava flows of the Kilauea series, but some of the fine grained, light colored ones may l)e from the pre-Kilauea series, which underlies the volcano. The common occurrence of course, chrysophyric rocks similar to that of the Uwekahuna intrusive lodly suggests that the throat of the volcano may l)e l)ordered by numerous intrusives like those in the walls of 1 —alemaumnau. G(ENERAI, PETROGRAPIY OF 14.ILAUEIA The lavas of Kilauea are basalts or picrite basalts with possibly some basaltic andesites in the pre-Kilauea series. Some of Noble's specimens of pre-Pahala flows were classified as andesites by Cross (38, p. I20). But an analysis of the predominant lava type of the pre-Pahala corresponds to a basalt, and slecimens collected by me from the pre-Kilauea are also basalts. None of the analyzed rocks of Mauna Loa or Kilauea is an andesite. Labradorite basalt, mostly with about one per cent of olivine is by far the most abundant variety, but there are some olivine basalts, and a few flows. Many of the gabbroid explosion blocks are picrite basalts containing 15 to 40 per cent of olivine. Only a few minerals and those of a rather constant character enter into the composition of the lavas. Olivine is l)resent as phenocrysts in nearly every specimen. It incloses small crystals of magnetite and round blels of glass. The interference figure shown by sections l)erl)endicular to an optic axis is nearly a straight bar, indicating an optic axial angle near 90~, which corresponds with a content of ferrous oxide of al)out 12 per cent. An analysis by Steiger (9, p. 295) of olivine from the MIauna Loa flow of 1852 gave 11.44 per cent of ferrous oxide. Pale brownish augite is the most abundant mineral, making up as much as 54 per cent of the rock. Small augite phenocrysts, though sparse and mostly poorly formed, can ble found in many specimens. Most of the augite occurs as irregular grains in the groundmass. Its optic angle as determined roughly in thin sections seems to be a little less than the normal angle (58-6o0). An augite from Haleakala, Maui was analyzed and described by Washington and Merwin (39). According to their interlretation it is essentially hedenbergite-diopside with small atmunts of acmite, clinoenstatite, and alumina in solid solution. It contains 1.89 per cent of titania and has an optic angle of 6i-62~ for red light and of 58-6o~ for )lue light. Hypersthene of a weakly pleochroic variety was found in only two specimens, b)oth of blocks ejected from Tlalemaumau, but hypersthene has been found by Cross in some of the pre-Pahala flows of MIauna Loa. In one of 22 Berncice P. Bishop Museu n-Bulletin 33 my specimens the hypersthene occurs as rounded phenocrysts and in the other it forms ragged ophitic plates inclosing laths of labradorite. Plagioclase feldspar in all specimens is a calcic variety, mostly a calcic labradorite with a composition near Ab1An2, but in a few specimens the phenocrysts are a sodic bytownite. Feldspar phenocrysts are practically universal. Magnetite and ilmenite appear in all specimens, usually as late products of crystallization. The magnetite forms beautiful crystallites in some glassy lavas. Brown glass is the only other common constituent. In a few nearly holocrystalline specimens it is full of tiny needles, probably of spatite. The glassy residue in many flows is opaque with iron ores. Cristobalite, recorded for the first time from the island of Hawaii, occurs in the vesicles of some of the coarser rocks. Sulphur and various sulphates have been formed in places by fumarolic action. The texture of the more completely crystallized lavas ranges from intersertal to diabasic, but the common material from thin flows or the tops of the thick flows has a black, glassy base. Stonc-Kila tca 23 THE EJECTED PRODUCTS CLASSIFICATION The materials ejected by Hawaiian volcanoes may be classified as rock fragments, ashes, bombs, and scoriae. Rock fragments are angular pieces of rock shattered by explosions, and range in size from those 2 millimeters in diameter to those with a greatest dimension of 3 or 4 meters. They may be classified according to size after the following scheme, which is slightly modified from Wentworth (40).' Maximum Size Minimum Size........... 256 mm. Bowlders or blocks 256 mm. 64 mm. Cobbles 64 mm. 4 mm. Pebbles 4 mm. 2 mm. Lapilli Volcanic ash includes all fragments of rocks, minerals, and glass less than 2 millimeters in diameter. The finest material may be called "dust." The term "ash" is also used in this paper in a general sense as including all bedded volcanic ejecta. The term "bomb" is restricted here to objects consisting in part at least of lava ejected in a liquid or semiliquid condition. As shown by Perret (30), the usual type of Kilauean bomb is a rock fragment with a coating of lava. Scoriae are cindery lava particles of varying size, which were ejected in a molten condition. It is not always possible to tell whether scoriae were formed contemporaneously with the explosion or whether they are older. Unusual glassy ejecta at Kilauea are Pele's hair, glass droplets or Pele's tears, and thread-lace scoriae. ASEIS IN THE PRE-IKILAUEA SERIES Several beds of fine grained yellow or reddish-yellow stratified ash are interbedded with the lavas of the pre-Kilauea series and exposed with them in the cliffs of southern Puna and Kau. The thickest beds observed are between 15 and 20 feet thick but pinch and swell abruptly. Five beds can be seen in the seaward face of Puu Kapukapu; but, because of the dis5Wentworth would restrict the terms "bowlder," "cobble," and "pebble" to rounded rock particles. In speaking of angular particles "block" may be substituted for "bowlder," but there seems to be no words in the English language to designate angular stones of the size of cobbles and pebbles. Until such terms are proposed and accepted, "bowlder," "cobble," and "pebble" must be used in a more general sense. 24 4Becrnice P. Bishop AlMuseumn.-Bulletin 33 continuity of the beds, there is no close agreement in the sections in different places, although the greater part of the ash is everywhere in the upper part of the cliffs. These beds agree in character and thickness with the pre-Pahala ashes of the Mlauna,oa slope as described by Noble (28, p. I19). The ashes in rainy sections of \Mauna Loa are decomposed, locally to clays, but the beds in the arid country south of Kilauea are quite fresh. THE PAHALA ASH Several areas along the Mauna Loa slope north and northwest of Kilauea are covered by deep soil and contrast sharply with the usual thinlycovered lava flows. One small area of this kind is on the small pali north of the Keauhou Ranch, another is the Bird Park kipuka, and there are several others, for instance those near the Kapapala Ranch gate. Such kipukas are made especially noticeable by the more luxuriant growth of vegetation on them. The sugar plantations around Pahala are supported l)y soilcovered areas of this kind. Tlhe Pahala soil, first mentioned by Dutton (12, pp. 97-98), was considered ly him an alluvial deposit formed on plains near sea-level. In I887 Hitchcock (i8, p. 55) advanced the theory that the soil was a decomposed volcanic ash, and later workers have accepted this interpretation. The ordinary surface material of the Pahala ash is a light, loose, ocher-yellow soil, similar in appearance to very fine sawdust. Hithcock says (17, p. 153) that the natives used to amuse themselves by jumping into banks of the ash just as children jump into snowdrifts. The loose surface soil, entirely fine grained and giving no convincing proof of its origin is in deeper exposures unmlistakably volcanic ash. In the dry stream channel near the northernmost tank on the Peter Lee Road the following section is exposed: Feet Loose, ocher-yellow ash................................... —........ 13 Stratified, indurated, gray ash........................................ 15 Base not exposed The yellow ash layer contains several thin bands of fine lapilli and gray ash and is pisolitic near the base. The lower layer consists of thin beds of gray ash (some of it l)isolitic) and lapilli. It is sufficiently indurated to form blocks in the stream bed. Examination of specimens of the Pahala ash with the microscol)e shows that it consists of fragments of yellow glass and a smaller amount of mineral fragments. Many of the glass fragments were derived from vesicular material. Stone-Kiilauca 25 The thickest section found is in the scarp back of the Kapapala halfway house, where about 95 feet of ash is exposed in a steep-sided gulch. The apparent thickness of this section may l)e in small part due to normal faults of slight displacement such as are exposed at nearlby places. Noble gives the maximum thickness of the Pahala ash as 75 feet. A number of exposures show plainly that the Pahala ash along the Alauna Loa slope is the product of two or more eruptions. In the section on Peter Lee Road is a sharp division between the two parts, the lower layer having a thin, reddened zone at its top. At some neighbloring localities the yellow layer alone is exposed and is underlain by an aa flow, which must be intercalated between the two layers. In a short valley just north of the Kau-Volcano road at an elevation of 2570 feet the following section is exposed: Feet Loose, yellow ash................................... —.............. 15 W ell bedded ash, mostly gray............-..-.................. 5 Slight unconform ity.................................................. Light yellow ash, contains pebbles and seems to be reworked.. 2 Thin bedded, yellow and gray ash....-.................................. 4 Base not exposed. At several places lava flows are interbedded with the ashes. There are also deposits of yellow ash in southeastern Kau. The flat-toppled hill, Puu Kaone (Sand Hill), is covered by yellow soil, which must be very fertile, for grass continues to grow upon it in spite of the lack of rain and the large number of goats that pasture uponl it. A small gully exposes 45 feet of ash made up of alternating yellow and gray ash. The yellow ash is fine grained but contains many lapilli and small )el)l)les; the gray ash, which is coarser, is in beds as much as i foot thick, and makes up about one-third of the section. Puu Kapukapu is a prominent hill capped bly 20 feet of yellow ash, and there is also a patch of ash on the Kulalauula pali. At least 6o feet of thin bedded, yellow and gray ash like that of Puu Kaone is exposed in a straight-sided landslide cirque southwest of the Keana Bihopa Kipuka. The same thick bed is exposed in several places in the tIilina pali at the top of the pre-Kilauea series. Because of its lithologic likeness and similar stratigraphic position the vellow ash of eastern Kau is correlated with the Pahala ash of the Maauna I,oa slope. Excavations on the Hilo-Volcano road near -Glenwood expose about IO feet of red clayey material, which is apparently much weathered ash. The deposit is unstratified except for thin carbonaceous streaks. The red color and clayey texture are believed to be due to weathering under lateritic conditions. (The annual rainfall at Glenwood is 228 inches.) The up)per part of the Pahala ash is red when freshly exposed in the rainy forest 26 Bcrnice P. Bishop Iliusiiscut-Bllcttiln 33 north of Kilauea, but when allowed to dry a short time in the sun, it takes on its characteristic ocher-yellow color. The Glenwood ash is here correlated with the Pahala ash. In addition to the lelposits in Kau and along the Volcano Road, yellow ash is found all the way to Ka Lae, the south point of Hawaii. and up Mauna Ioa to an elevation of 7,500 feet. Ash deposits are also found in other parts of Hawaii, especially along the Hilo and Hamakua coasts, where they support large sugar plantations, but where there is no evidence that the deposits had a common source with the yellow ashes of Kau. Within the Pahala ash itself are breaks but no complete changes in size or character of material, so that it is justifiable to assume that it came from a single source or at least from closely related sources. Several possible sources have been mentioned. IEmerson (14, p. 435) suggested that the so-called "caldera" of MIohokea was a logical source, inasmuch as it lies in the midst of the region where the ash abounds. Recent work, however, has shown that Mlohokea is not a crater but only a great valley formed by faulting and erosion. IIitchcock regarded Mokuaweoweo, the summit crater of Mauna Ioa, as the most lrol)al)le source for the Pahala ash because he had noted ash deposits, which he considered equivalent, at various localities all around Mauna Loa. The absence of the deposit around the summit crater he explained by the assumption that the ash was carried so far into the air and the force of the eruption was so great that no ash fell near the source. This reasoning is l)roved false by the proximity to Kilauea of the ashes from its erulptions. No ash beds occur in the walls of Mokuaweoweo, so that were Maunna,oa the source of the yellow ash, the summit of Mauna Loa must be post-Pahala. In another connection Hitchcock (17, P. 148) suggests that Puu o Keokeo on the southwest flank of Mauna Loa may have been the source of the Pahala ash, but the text does not make clear whether the suggestion is to be regarded as the author's own or that of Dr. S. E. Bishop. The summit cones of Mauna 1Kea have been often cited as evidence that that volcano closed its active history with great explosive ash eruptions. If this view is correct, Mauna Kea might be considered as the main source of the Hawaiian ash deposits. But examination of these cones show s that they are not ash cones, as has often been supposed, but cinder cones quite like those at the sources of many Mauna I,oa flows. It is true, however, that the eruptions of Mautna Iea were more explosive than those of MIauna Loa, for they built bigger cones and producedl many more bom)bs and some beds of ash. The headward gullies of the Wailuku River and other dry stream channels expose a few ash beds as much as 10 or 15 feet thick inter Stone-Kiloa 27 27 bedded with the uppermost lava flows, but there is no great surface mantle of ash as there would be if Maunna Kea were the source of the ashes of the whole island. MAany of the cones of Mauna Kea are above the trade wind zone, as was pointed out to me by Finch, so that the great amounts of ash in Rau cannot le explained by wind tralslportation from Mauna Kea. It is hardly conceivalble that lalpilli could have lbeen lblown from Maunna Kea to Pun lKaone. E. D. Baldwin (3) suggested that... at some an eiit periol there has been a great line of yellow erlulpmons, extending from Pun lkapukapu.... past the I-amakaja hills to the lower portion of hian, and that the sources of this yellow ernption in the lower part of Kani have been covered up with later flows.... and that the great beds of yellow soil that wVC find tolay all over Kau were blown there fromt these sources. The cones whose yellow "tufa" reminded Baldwin of the Pahala ash are, however, cinder or lava cones, and show no evidence of having prolucel ash. Is it possible that Kilauea was the source of the Pahala ash? EmJierson (14) suggested this idea only to reject it as "extremely improbable." Nev~ertheless, there is considerable evidence in its favor. With the exception of the Pahala ash, the thickest deposit of aslh kniowni onm I awaii is at IRilauea, so there is no questioning the potency of IKilauea as an ash-producer. MNoreover the distribution of the Pahala ash is exactly what would be cxIectel if Ibilanuea were its source and the trade winds the tranisporting agent. It is similar to the distribution of the IKilauean ashes. Baldwin pointed out that the (istrilbution agreed with his theory of a source in the region of the Kamakaia H1ills, but it agrees even better with the theory that Kilauea was the source; for the greatest thickness found is north of Kamakaia. 'I'he miscroscopic character of the Pahala ash is the same as that of the older Kilauean deposits. The ash at Puu Kaone is somewhat coarser than that near Pahala; this condition can he exlplainted if Klilauea were the source, but not if the material came from Mauna Loa or Mauna K-ea. The (Glenwood1 asi, now so thoroughly decomposed, was p)roblably tine grained, as would be exlpected on the windward side of the source. The source of the Pahala ash can be letermined finally only by a stuly of the ash (lelposits of the whole islan d. ~A SI I ES) 01F K-lAU I," C RA TIR,r i, r UWF-.KA1.1t'-N.,\ Asii The only ash bed in the walls of IKilauea is exposed at the lbase of the ixekahuna cliff northeast of the fault terraces. Another ash section was formerly exposed southwest of the fault terraces (31, p). 230) but was buried by the lava flow of 1919 from Halemanimau. A careful examinatiion 28 Bernlice P. Bishop Aluseums-Bulletin 33 of the walls showed no other beds, but ash layers only a few inches thick might escape observation. The exposed ash bed extends along the base of the cliff for about 3,000 feet from the niche at the northeast end of the fault terraces to the point where it (lips below the level of the crater floor. At the northeast end of the exposure the bed is three feet thick and consists of gravel, a few lavacoated bombs, and rock fragments commonly 6 or 8 inches across but reaching a maximum size of about I6 inches. Beneath the Uwekahuna intrusive body is 6/2 feet of ash having a basal layer of medium-coarse black ash overlain by 2 feet of gravel and cobbles, then another ash layer overlain by a second gravel stratum, and at the top about 8 inches of yellow ash with glass droplets and crushed thread-lace scoriae (II, pp. I63-I66). At the southwest end of the exposure the ash bed lies on the surface of an unconformity in the lavas. According to Powers, the incomplete section formerly exposed at the other locality was 17 feet thick and was composed of yellow ash containing rock fragments i or 2 inches across. THIE OLDI)R SURFLACE ASHEIS The surface around IKilauea is covered by a deposit of volcanic ash, which is thick enough to be noticeable (at least i inch) on the Hilo road six miles south of Kilauea, beneath the Keiwa flow along the Great Crack sixteen miles southwest of the volcano, and on the Mauna Loa slope in Ohaikea four miles to the west. The ash deposits have a maximum thickness of 35 feet in the cliffs forming the southeast rim of Kilauea crater. The thicknesses at other places around the crater are: 4 2 feet on the top of Uwekahuna, 4'2 to 6 feet at the Observatory, 2 to 32 feet on the road east of Kilauea Iki, and 13'2 feet on the northwest rim of Keanakakoi. (See fig. 3.) The floor of the crater is covered b)y recent flows so that the only ash on its surface is that of the explosions of I924. The ash deposits vary from pllace to place not only in thickness but also in the proportions and coarseness of the different materials of which they consist. A bed of the remarkable light pumice, known to the Hawaiians as "limu" (moss) and called "thread-lace scoria" by Dana, forms a conspicuous basal layer of the ash deposits around the northern rim of the crater, where sections were especially well exposed during the summer of 1925 by work on the roads. "Thread-lace scoria" is a very expressive name for a light greenish-yellow froth of basaltic glass, whose vesicle walls are reduced to a skeleton of glass threads so that it can be easily crushed between the fingers. D)ana's description leaves little to be added. Stonc —Kilauca( 29 The bed of thread-lace scoriae lies directly upon the topmost lava flows nearly everywhere, but along the road between the Volcano House and Kilauea Iki a thin, discontinuous bed of purplish-gray, clayey ash underlies it locally. The scoria bed itself reaches a thickness of 2/2 feet in pockets but is less than i foot thick in most places. Figure 4 shows two areas of greater thickness, one north of Kilauea Iki and a larger one north of the main crater. Had Kilauea Iki been the source, there should be thick deposits of scoriae on 2.r 7. a -I. - j.Z. 27.'6..., **0*'...s. 2 -6~ /. iI.-$..V0 '1.0 0 0i: I C." ^ 0c' IM, t FIcURE 3. Map showing the distribution of the surface ashes of Kilauea. Thicknesses in feet and inches. Roman numerals show the location of the sections given. in the text. its inner walls, none of which was found. The scoriae were evidently produced by fountaining in the main crater, probably in the northern and northeastern portions, during a period of southwesterly, or "Kona" winds. A few days would have been sufficient for the formation of the scoriae. Thread-lace scoriae, besides occurring in the bedl just descriled, are also found in the Uwekahuna ash and in pockets beneath the ash of 1924 around the southern part of Kilauea crater. The latter deposit locally reaches a thickness of several feet but consists of a coarser variety of scoriae than the basal layer. In the younger material some of the vesicles have complete walls of glass, and many of the scoriae are inclosed in a thin, brown glass 30 Berntice P. Bishop Museuini-Bitlietin 33 membrane. Identical scoriae occur around many of the Mauna boa sourcecones and also in the Kamakaia cones, but at only a few places does it have the perfection of the scoriae of the basal layer. iI', z'. J. z.2 y -.7 /2.6 1. /Z - I.1 N I Tr-ee.3 7 7.4 A P - 4, —, / e. 3O K e n k t ko 0 IMrl'! FIT(URFt 4. Mlap showing the distribution of the basal laYer of thread-lace scoriae around 1K ilauca. Thicknesses in inches. Overlying the thread-lace scoriae around the northern end of Kilauea are well bedded ash dleposits. Typical sections, whose location is shown in figure 3, are as follows: Stone-Kilauca 31 I. Feet Itches Coarse gray ash with pebbles and lapilli......................................... 9 Tan ash.........-......................................................... o I Coarse well-bedded dark-gray ash with a few lapilli............................. o 9 L apilli........... -..-..-........................................... o o,-i Coarse black ash............................................................... o 6/ Black to greenish-gray ash.-................................... —. Ii/, Tan ash with red top and thread-lace scoriae at base........................... o 2 Dark-gray ash................................................ - 8 Thread-lace scoriae....................-.................................. o Ir Clayey gray ash.................................................................... 5r P a ho eh o e................................................................................. 6 6 IT. Feet Inches Ash with streaks of lapilli and gravelly top........-....................... 4 Yellow ash........-........................... o 21 2 Gray ash.......................-.................... o 4 Black scoriae........ —....... —... -... -............ -.... --- —-.. o I Gray ash..................................................... o 3 Ash with carbonaceous material at top.............................-........ o 4 Clayey gray ash with red top..............................-.....-........... o 2 Thread-lace scoriae -....................................... o 5 P ah o eh o e..................................................................... 3 I2 III. Feet Inches Coarse gray ash with scoriae at base....-........................... r1 9 Gray, clayey ash.................................... o 4 Thread-lace scoriae........................ -- ------------ - ---------------—. 3 Fresh pahoehoe... --- —---.. --- —---------------------- -—........ 3 4 IV. Feet Inches Ash and lapilli.................... --- —-- -.... -—. --- —. --- —-----—.. --- 0 4' Fine, bedded ash.............................. --- ——....... o 8 /2 Fine, clayey ash with dark top —... —.. --- —. —. —. --- —---------—. --- ——.. --- —.. o 2/2 Thread-lace scoriae........... -..-......... --- —-------—...- --—.. --- —-... ------—. o 7 Pahoehoe............ --—. ---------- ----—. --- -------------- --------—. ---- ---------- -—... I Io/ The ash beds thicken abruptly south of Uwekahuna on the west side of Kilauea and Keanakakoi on the east, but they are strictly equivalent to the ashes north of the crater. Traces of the basal layer of thread-lace scoriae can )e found at the base of the ash section in the cliff at the east end of 32 Bcernice P. Bishop AlItuscum-Bulectin1 33 the flow of 1921, and the deposits are continuously exposed around the rim of the crater. Two sections are: V. Feet Inches Powlder gravel....................................... 6 Ash, upper part coarse gravel and cobbles, scattered cobbles in lower part --.................................. 5 o1 Black scoriae....-.. —............... ---------------------------------------—............ o Compact yellow and gray ash.............................-..-....... 9 Crushed thread-lace scoriae..................-...-.....-............... o 3 Pahoehoe....... --------------------------------- —........ — ----------- --—... -- -- 18 2 VI Bowlder gravel................................-... o Bedded yellow ash.....-.........-................................. 9 o Black scoriae............-....................................................... o 6 Yellow ash....................-............... 22 o Thread-lace scoriac....-................................... o 3 33 9 The bowlder gravel, which consists of rock fragments of all sizes up to blocks weighing several tons, and which has a maximum thickness of ten feet in the sand spit near Halemaumau, lies disconformably on the lower beds. EInough fine ash accompanied the bowlder gravel to give it a smooth surface, over which are scattered large bowlders thrown out in the final paroxysms of the eruption. Many of these bowlders are large; a huge one weighing about eight tons lies 6,000 feet southeast of the present pit, and another weighing about two tons was seen just west of the summit of Uwekahuna, a mile north of Halemaumau. Scattered bowlders can be found as far away as the Keauhou Ranch, and lava-coated bombs, some as large as a man's head, are locally mingled with them. Impact pits formed in the fine surface ash by the fall of the larger bowlders and bombs are still Ireserved. The bowlders either remained in their impact pits or bounced on, many of them breaking on striking the earth. Good sections of the ashes are exposed along the cracks, which extend southwest from the rim of the crater. A section near the crater is: Stoi;e-Kilaucea 33 VIT. Feet Inches Fine, gray ash of the eruption of 1924....... ----.................... o 5 Coarse thread-lace scoriae, many with black, glassy skills............. 3 Fairly well bedded, gray ash with pebbles..........-................... 8 Pebbles and cobbles............-......................................... —................ — o 8 Brown ash containing pisolites and pebbles.............................. o 6 Coarse gravelly ash........................................ 1 3 D isconform ity ---............................................. lMedium coarse, grayish-yellow ash with pisolites.-.........-.........-.... 2 o Very fine, dark brown pisolitic ash......................... ()o G rayish-yellow ash............................................................. 4 10 Coarse black ash..-..................................-...... 4 2 P a h o e h o e.-......... --- — ---- - -- - - - - -- -------------------------------- - ---- --------------—... -... -—........ 19 5 The ash deposit thins abruptly southwestward, five miles from the crater being less than one foot thick. The ash near Mauna Iki is in four layers: (i) a basal layer of compacted, fine, light yellow pisolitic ash, whose surface is sun-cracked; (2) loose, medium, greenish-black ash; (3) compacted, yellowish-gray pisolitic ash; and (4) a scattering of pebbles and thread-lace scoriae. Some of the pisolites in the third layer are three-quarters of an inch across. The finer ashes examined under the microscope were found to consist of mineral fragments and glass fragments. The glass fragments are l)arts of glassy scoriae coarser than the thread-lace variety, and some contain stretched vesicles. Some of the fine, light gray ash is more than half mineral fragments, but the black ash and especially the indurated yellow ash of sections No. V and No. VI are more than 75 per cent glass. The surface ashes, with the exception of those of I924, were deposited before the earliest written records of Kilauea, but Ellis gives native traditions of an eruption in I789 or 1790, which killed a number of people ( 3, P). 186-187). The story can be found repeated in nearly every account of the volcano. Direct evidence of a most unusual kind proves that there was an eruption witnessed by natives: in the spring of I920 R. H. Finch discovered fossil hunan footprints in the pisolitic ash of the Kau desert. Other footprints have since been found in surprising numbers at many places along the route of the ancient trail to Uwekahuna (23, pp. 114 -ii8, 156-157). Unmistakable footprints are found in 1)oth the ulpper and the lower pisolitic layers, so that they were unquestionally formed during the mud-rains that deposited the ash. (See Pl. I, B.) The surface ash near Kilauea was gravelly and unsuited to retain the impressions. Dana considered the surface ash as the product of the eruption of 1790, but Hitchcock thought it imnlroal)alle that all the ash was of one age, and 34 Bertlnicc P. Bishop luseuml-Bulletiul 33 Powers (31, p). 232) mentioned the disconformity beneath the gravelly ash near Halemaumau, which proves that there were at least two eruptions. In the present paper the coarse upper layers of ash all around the crater together with the coarse gravel near Halemaumau and the footprint layers of the lau desert are considered to be products of the eruption of 1790. The ashes older than 1790 are not, however, the products of one eruption. The ash sections around the northeastern part of the crater show black, carbonaceous layers mostly about half an inch thick and contain remains and impressions of ferns and other plants. The ash immediately below these soil layers (for such they are) is reddened in places. The soil layers are somewhat discontinuous, though four intervals between eruptions are shown quite definitely. The basal member of the ash series is a thin layer of gray, clayey ash, which occurs only locally and whose existence as a separate unit is doubtful. Next above the gray clay comes the layer of thread-lace scoriae and a small amount of associated ash, which is the basal layer in most places. Between the thread-lace scoriae and the coarse ash of 1790 are two thin ash deposits separated by a soil layer, and another soil layer overlies the upper one. Including the ash of 1790 there is thus evidence of four and possibly five periods of explosive eruption in prehistoric times. It is proballe that there were more eruptions which, like the eruption of 1924, left no trace around the windward side of the crater. The yellow tuff along the southern rim of the crater corresponds to the ash of the two eruptions next preceding 1790, for traces of thread-lace scoriae occur at its l)ase, and the ash of 1790 overlies it. The lee side of the crater, however, is a desert, so that no plants grow to mark the intervals; and it cannot ble determined how much of the ash is due to any one eruption or whether there were several eruptions. There is some evidence that the layer of l)lack scoriae in sections No. V and No. VI (p. 32) occurs at or near the top of the lower deposit. An ash bed at least three feet thick shown in an open fissure near Cone Peak is separated from the surface ash by 12 feet of lava flow, which seems to be a flow from Cone Peak. No other ash sections could be found in neighboring fissures, but the field relations suggest at least that the Cone Peak lava flow occurred during the period of explosive eruptions. The age in years of the ash beds older than those of 1790 can only be estimated roughly. The lava flows below the ash are fresh, and the soil layers between the different deposits are thinner than the post-1790 layer at the surface. The establishment of a forest, which has taken place entirely since the lava flow on which the ash lies, is apparently not yet complete, for all the trees are young, and old inhalitants have noticed an advance of the forest within the last o3 years. The largest tree I saw near Kilauea was Stonce-Kiltltca 35 a five-foot koa near the Hilo road at an elevation of 3,800 feet. Dr. 1. IB. H1. Brown of the Bernice P. Bishop Museum estimated from data furnished by me that the tree was between 150 and 300 years old, prol)albly nearer 300. This tree is quite surely older than the ash of 1790. The age of the oldest ash is prol)al)ly not over 300 to 500 years. TH: Asii o0 1924 Kilauea was erupting explosively from May o1 to May 24, 1924 (26), and during this period threw fragmental material over a large area. The amount of ash ejected was far less than in 1790 for the maximum thickness of the ash of 1924 is about I12 feet compared with m(ore than o1 feet in 1790. The floor of the crater for at least i,ooo feet from the pit is thickly covered by angular rocks. The largest lies near the east rim of Halemaumau, is T feet long, and has an estimated weight of 14 tons. (ne block weighing 8 tons was thrown 3,500 feet, and several were thrown nearly 5,ooo feet. The blocks are nearly all freshly broken and show no sign of fusion although many were incandescent when ejected. H eat crackiing may have been partly resl)onsille for the breaking of many of the blocks when they fell. No bombs have been found. The light gray ash formed by the explosion, fell either as dust or as pisolitic mud rains. Fine dust was carried 25 miles or more to the southwest. Along the southwest rim of the crater the ash is 3 to 5 inches thick. EIxamination with the microscope showed that the ash consisted almost entirely of mineral fragments and that the few glass fragments were l)rol)ably pieces of older rocks rather than of fresh lava. 36 Bernlice P. Bishop iluitscu l-B ulletil 33 ST R UCTURE TIHE KILAUEA DOIMlE Tol)ogral)hically ilauea is an independent dome on the flank of Mauna Loa, but because of the very gentle slope of both Mauna Loa and Kilauea its dome-like nature is not apparent in some parts of the area. In fact, before the publication of the topographic map, some geologists considered that Kilauea was merely a sink in the side of Mauna Loa. The topographic map shows, however, that the only place where the slope is such that flows from Mauna Loa might reach the crater of Kilauea is a strip somewhat less than one mile wide along the foot of the Mauna Loa slope north of Keauhou Ranch. A flow reaching the Kilauea dome to the west of this strip would be deflected off to the southwest, as was the Keamoku flow; one to the east would flow off to the northeast. As most of the critical slope is covered by the Pahala ash, it is evident that no great amount of recent lava has come down there. Under these conditions it would have been impossil)le for flows from Mauna Loa to build the Kilauea dome. Two explanations of the Kilauea dome seem possible: either it has been built up by the volcano itself; or it has been arched up by pressure from l)elow, for instance by a laccolithic intrusion, as Daly (io, pp). 109-Ii) postulated. Several lines of evidence, however, show that the lavas really came from the vicinity of the present crater: I. In several kipukas on the flanks of Mauna Loa along the KauVolcano road and as close to Kilauea as the Bird Park, there are thick deposits of the Pahala ash interledded with the Mauna Loa lavas, yet the Pahala ash does not appear at Kilauea either at the surface or in the 450 foot section at Uwekahuna. The flows of the crater walls cannot then have come from Mauna Loa. 2. Many of the long flows of Mauna boa are aa like the Keamoku flow near Kilauea, but aa flows do not occur in the walls of Kilauea. 3. In the Uwekahuna ash interbedded with the lavas of the crater wall are ejected bowlders 3 or 4 feet across as well as large lava —coated bomls, which obviously came from a near-by source. Therefore, there must have been a volcanic vent in the vicinity of the present Halemaumau before the wall of Uwekahuna was built. 4. On the lack slope of Uwekahuna the ropy p)ahoehoe surfaces indicate a direction of flow to the west, so the slope has not been reversed Stone-Kilaicea 37 by tilting. Because of the heavy cover of ash, exposures are comparatively few. 5. A final bit of evidence is the homology with Mokuaweoweo, the summit crater of Mauna Loa. The similarity in all essentials between MIokuaweoweo and Kilauea is one of the most striking impressions made on an observer, and in the case of Mokuaweoweo there is of course no doubt as to the source of the lavas of its walls. The Kilauea dome was therefore built up, probally by the co-operation of two processes; overflow from a central lake and fissure eruption. Flows from Halemaumau have built a lava cone in the present crater and in 192I overflowed the crater wall for a short distance. Fissure eruptions have occurred frequently southwest of the crater. The way the contour lines bend around the eastern end of Kilauea Iki, which is the highest point near Kilauea with the exception of Uwekahuna, suggests that large quantities of lava were extruded there. Road cuts in the vicinity show thin, vesicular flows such as occur close to the sources of some eruptions. The Twin Craters were undoultedly the source of some lava. THE KILAUEA-MAUNA LOA CONTACT The relation of Kilauea to Mauna,oa never has been satisfactorily determined. Dana (1l, pp. 260-264) believed at first that Kilauea was younger than Mauna Loa and subsidiary to it, but his later reasoning inclined him to believe it independent and perhaps younger. Dutton (12, pp. 120-I21) insisted that Kilauea was an independent volcano, whose lavas had interfingered with those of Mauna Loa. He did not make any statement as to their relative ages. In recent years Jaggar (20, p. 198) has held that Kilauea is much older than Mauna Loa, whereas both Daly (Io) and Noble (28, p. 121) have believed that it is only a sink in a platform of Mauna Loa lavas. None of these conclusions, however, was based on detailed field work around Kilauea. A study of the well-marked topographic boundary between the Mauna Ioa and Kilauea domes shows important structural relations. On the northwest the Mauna Loa slope descends in a series of terraces, which were noticed by Dutton and regarded as of alluvial origin and so as evidence of uplift. This interpretation was largely due to Dutton's failure to recognize the Pahala ash as ash. The terraces are plainly abnormal features of the topography that cannot be explained by any vagaries of stream erosion or lava flows but must be either wave cut terraces or fault terraces. As the terraces are covered nearly everywhere by post-Pahala lavas, which have cascaded over them, and as vegetation is fairly heavy, direct observation is difficult. An examination of the topographic map A A, -, - '- -.-. - M-. A -A FIGURE 5. Structure sections across the Kilauea dome. A-A is drawn across the Puna ridge through the crater of Kane Nui o Hamo and the eastern pit of Makaopuhi. A line drawn from the base of the Kilauea series at the north edge of the area to the approximate base of the series in the fault-cliff south of Makaopuhi does not quite pass below the bottom of the western pit of Makaopuhi (not in the line of the section). As the walls of Makaopuhi are entirely made of the Kilauea series, this section shows that the Puna ridge has been built up by flows of the Kilauea series. B-B shows the crater of Kilauea with the agglomerate-filled throat of Halemaumau and also the small inward-facing cliffs southeast of Kilauea. C-C shows the fault-cliffs along the sea coast and the shallow graben of the southwest rift zone, also the hypothetical graben in the pre-Kilauea series. D-D shows Puu Kapukapu, a horst of preKilauea lavas capped by the Pahala ash. (All profiles drawn from the photolithic proof sheets of the United States Geological Survey Topographic Atlas on the scale I:31680 and exaggerated vertically about twice.) Stone-Kilautca 39 gives no support to the hypothesis of marine origin, for the essential feature of wave cut terraces is their horizontality, but the Mauna Loa terraces slope to the southwest, one of the prominent ones about 900 feet in 8 miles. Profiles across the terraces show that the slope on their "treads" is no less than that above or below the terrace. (See fig. 5, section D-D.) There is not the slightest trace of up-raised beaches or reefs nor are there any signs of uplift in other parts of Hawaii in Kohala or on the slopes of Mauna Kea, where there have been no recent flows to mask them. In addition to this negative evidence there is positive evidence of faulting. In a few places truncated flows and ash beds are exposed under the cascading flows, and along the tops of many of the palis there are open fissures in the recent flows, showing a repetition of movement along the old lines. (Such fissures might, however, be due merely to incipient landslides.) About one mile north of the Kapapala Ranch gate, beside the Volcano road is an area about three-quarters of a mile long which has not been covered by recent flows. At the upper end of this area at an elevation of 2I00 feet is a narrow horst composed of horizontal, truncated flows and ash beds; at the lower end of the area is a long, closed depression, which is on the continuation of the faults forming the horst. This closed depression could only have been formed by faulting. Another small horst about oo00 feet wide lies parallel to the road at an elevation of 2215 feet. A row of holes along its western edge marks the course of a fault. All these faults and terraces are exactly parallel to the series of fissures making up the southwestern rift system of Kilauea. From Kapapala Ranch southwest to Honuapo, a moderate slope about two miles wide intervenes between the Kilauea lava field and the foot of a steep terrace, whose front edge rises from about i8oo feet at Honuapo to about 2600 feet near Wood Valley. At Honuapo the terrace swings around to the west past Naalehu. Mohokea and Wood Valley are great amphitheaters in the face of this terrace. Wood Valley is a flat-bottomed, steepsided depression a mile wide and two miles long, whose form suggests an origin by faulting and landslide sapping. It may be significant that the sides of the valley are either parallel to the cross-faults of Puu Nahaha or are actual continuations of them. Mohokea is five miles across and five or six miles long and resembles Wood Valley except that it has several large hills or buttes inside its walls. The manner of origin of Mohokea is doubtful, but there is clear evidence of faulting across the valley. It appears, then, that circumferential faults of large throw were formed along the Mauna Loa slope previous to the most recent lava flows. The exact age relation of the faulting to the Pahala ash is difficult to determine, 40 Bciricc P. Bishop It lStun1f191B(llCfl ci'B one reason being that it is apparently not the product of any one eruption but may have l)een deposited over a long period of time with considerable pauses intervening between eruptions. Good exposures are rare, but in one locality an ash led a few feet thick continues down the slope of a terrace between two lava flows. The very thick bed of ash back of the halfway house, however, is cut by normal faults of a few feet displacement, and at several other places there is plain evidence of faulting in the Pahala series. Noble and Clark defined the Pahala series as a group of flows with interbedded ashes cascading over cliffs of the pre-Pahala lavas. They did not determine to what extent the Pahala ash was cut by faults. It seems proballe that the main faulting took place toward the end of the Pahala period and was completed before the first of the post-Pahala flows. The post-Pahala flows, which have cascaded down over the terraces, apparently have no great aggregate thickness, for in several places there are kipukas in them floored by the Pahala ash. Bird Park is a kipuka in the Keamoku flow about two miles from Kilauea, and its floor is covered by the Pahala ash. Pahoehoe flows from Kilauea cut across the foot of the kipuka and overlie the ash. Flows from Kilauea also abut against the steep, ash-covered slope where the Puu Oo trail reaches the foot of Mauna Loa. The Pahala ash is older than the surface flows from Kilauea at every place exposed. The Pahala ash does not appear in the walls of Kilauea crater and so is not only older than the surface flows but is older than all the flows exposed in the cliff at Uwekahuna. Apparently the Pahala ash lies on a slope against which the lavas of the Kilauea dome abut. If the Mauna Ioa slop e e projected toward Kilauea, it passes below the crater floor even if no allowance is made for possible buried terraces or for recenlt faulting in the IX Tv. Tl. 1 t St 1l }:'B | |; uk - - H S -; '<~~~~P"~ r< 1r~38 1;is~ i ' '> r. S 1z-. m Sgi.~at - *;;;M- N1 i~' 1 1li~:i: Il.\ii` ii; 1I1I jt~lkily;llw 1 Ii 1, 1Xl'~ ll i xsil iiW1 1/Xp11-r S/}t'IIg \ tSI( X11\1.. '11(>1( fx1 \1 [l13\'' \[X(;(' 31 I p r 0x i cr P I ~j i6,r M v N~'s, r, c I g, '5N i. li, a 3 | ",0 I I 01'1 N VI'IS, ~s. 1 ~ I ~1 it I,, A IF F; Sl VN; Sf0 I '''l I F F N ' FI' j;XI I i l \ I' 1; \F;1I "I ('(>N'VFR1V0 WITH1& XHt '1)<1 r'g111& j'llo't I o", I To renew ie, book must be brought to the desk. TWO WEEK BOOK DO NOT RETURN BOOKS ON SUNDAY DATE DUE t o ' Form 7079 7-51 30M S UNIVERSITY OF MICHIGAN 3 9015 02327 1946 AUG28 i944 UNIV. OF MICH. LIBRARY FLIBRilmed bv Preservation 1997 Oflg mall R H H ERE N