LIBRARY 
 
 UNIVERSITY OF CALIFORNIA. 
 
 GIFT OF 
 
 Cto 
 
IV. 
 
 LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 BY 
 
 WILLIAM H. PICKERING. 
 
 WITH SIXTEEN PLATES. 
 
 PRESENTED FEBRUARY 9, 1906. RECEIVED APRIL 7, 1906. 
 
MOKUAWEOWEO AT NlGHT. 
 
 Compare with Figure 18 taken at about the same date. 
 
 
LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 THE lunar surface presents such a strong contrast to the more thickly populated 
 portions of the Earth, that little resemblance between them can be traced. It has 
 therefore naturally proved very difficult to explain the nature and origin of many of 
 the features of our satellite. Even those of our volcanic regions which have been 
 most extensively studied, show little analogy to the Moon. There are other regions, 
 however, notably in the Hawaiian Islands, where an entirely different class of volcanic 
 phenomena are exhibited. These it is now found bear a striking resemblance in some, 
 respects to what we find upon our satellite. Although the Hawaiian craters are mostly 
 extinct, or at present inactive, yet they are the only ones known of this type exhibit- 
 ing any activity whatever. 
 
 In view of these facts the writer determined to visit the Hawaiian Islands in the 
 summer of 1905, and study their volcanic features with especial reference to those 
 found upon the Moon. In Hawaii a considerable number of the craters are of the 
 engulfment type, as distinguished from those of the explosive type, so well developed 
 in southern Europe. In the latter class a high truncated cone is built up by mild 
 eruptions of steam and cinders, sometimes alternating with lava. At long intervals 
 violent explosions occur, which sometimes blow away a large portion of the summit, 
 thus entirely changing the shape of the mountain. Such an explosion of steam 
 occurred in Vesuvius at the time of the destruction of Pompeii, and a still more violent 
 one in Krakatoa in 1883. Nothing whatever of this sort is found upon the Moon. In 
 volcanoes of the engulfment type on the other hand, comparatively little steam is 
 evolved, often there is no exterior cone, and the craters enlarge quietly by the crack- 
 ing off and falling in of their walls. 
 
 The Hawaiian structures, although similar to those of the Moon, are comparatively ' 
 on a very small scale, and their dimensions must often be multiplied by a factor of 
 from 100 in the case of the older craters, to 300 in the case of the more recent ones, 
 in order to equal the dimensions of the similar formations found upon our satellite. 
 This applies especially to horizontal distances, vertically a factor of from 
 20 more nearly represents the proper proportion. 
 
152 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 The force of gravitation at the surface of the Moon is but one- sixth as great as it is 
 upon the Earth, and this difference is usually given as the cause of the great compara- 
 tive size of the lunar formations. On the old theory that the lunar craters were due 
 to explosions of steam, like our explosive volcanoes, it was evident that matter 
 expelled from a crater vent could be thrown six times as far as upon the Earth. 
 Although this theory is now practically abandoned, gravitation would still have an 
 influence on the relative size, since a cliff or pinnacle upon the Moon could be six 
 times the height of one upon the Earth, and yet exert no greater crushing force on 
 the material beneath it. Still it is very evident that this explanation alone is inade- 
 quate to account for the great difference in size actually observed. 
 
 The facts seem to be that we are really trying to compare objects formed under 
 entirely different conditions. The larger craters on the Moon came into existence 
 when the thin, solid crust covering the molten interior was, owing to the solidifica- 
 tion and contraction of the crust, much too small to contain the liquid material. 
 The craters were therefore formed by the lava bursting through the crust, and 
 so relieving the pressure. 
 
 Later, after this relief had been found, and the crust had thickened, the interior 
 regions by cooling shrank away from the solid shell which was now too large, and 
 being insufficiently supported caved in, permitting the great fissure eruptions which 
 produced the tnaria. These extensive outflows of lava dissolved the original solid 
 shell wherever they came in contact with it much as they do at the present day in 
 Hawaii. Had the Moon been much smaller, these extensive eruptions would not 
 have attained such relatively great size, or might even not have occurred at all. On 
 the other hand, had the Moon been larger, their relative size would have been greater, 
 since the volume of the sphere would have been larger in proportion to its surface and 
 would therefore have shrunk more in proportion. This was precisely what took place 
 upon the Earth in all probability : our original gigantic craters were destroyed by the 
 outflow of the earlier archaic rocks, which completely submerged and dissolved them. 
 Our present Hawaiian craters must therefore be compared, not with the primary 
 formations still left upon our Moon, but rather with the secondary ones formed 
 later upon the surface of its maria. Of these Bessel, twelve miles in diameter, 
 is a large and well known example. From this size down countless craterlets are 
 known. 
 
 Three craters are found upon the Earth measuring about fifteen miles in diameter. 
 They occur in Kamchatka, in Japan, and in the Philippines, but are all of the explo- 
 sive type, and therefore not comparable to those found on the Moon. It is possible 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 153 
 
 that a large engulfment crater formerly existed upon Kauai, and another in southern 
 Hawaii, near the coast, south of Mauna Loa, but the writer was unable to examine 
 either of these regions during his recent visit. The latter crater must have been 
 about five miles in diameter, the former perhaps much larger. The largest engulf- 
 ment crater known is Crater Lake, Oregon, measuring five by six miles in diameter 
 with a depth of about 3,000 feet. Next to it comes Haleakala in the Island of 
 Maui, Hawaii, measuring seven miles in length by two in width. It is about 2,000 
 feet deep. 
 
 The secondary craters found upon the lunar maria are so small that it is impos- 
 sible to study their interiors to advantage; we shall therefore content ourselves 
 with comparing the Hawaiian formations, as far as possible with the large pri- 
 mary formations of the Moon, without regard to the great discrepancy in their 
 relative size. 
 
 On the Hawaiian Islands with the exception of the three groat craters of 
 Haleakala, Mokuaweoweo, and Kilauea, few of the crater pits exceed half a mile in 
 diameter, measured on their crater floors, or former free liquid lava surfaces, although 
 there are probably several hundred pits over 200 feet in diameter. In addition to 
 these are countless cinder cones, ^piracies, etc. On the Earth at present the cooling 
 process always intervenes before great size is attained. Doubtless formerly the lava 
 was hotter when it first issued from the interior than it is now, also the solid crust 
 resting on the liquid mass was thinner, so that the channel communicating with the 
 interior was shorter and of greater diameter, thus offering a freer passage to the 
 
 liquid flow. 
 
 Terrestrial craters may be divided into three classes, according to the materials of 
 which they are composed. These are (a) tuff or tufa cones, which are made of 
 hardened volcanic mud, (b) cinder cones, made of scoria, lapilli, or sand, that is, lava 
 broken up into masses of varying size, by the action of steam, from stones several 
 inches or even feet in diameter to fine powder, and (c) lava craters, where the lava 
 occurs in unbroken masses. It is this third class, where less water is involved in the 
 eruption, which most resembles what we find upon the Moon. Representatives of all 
 three classes are to be found in Hawaii. Many volcanoes like Vesuvius eject both 
 
 cinders and lava. 
 
 The third class may again be divided into four subdivisions according to the 
 of the craters, namely: lava cones, lava pits,^a rings, and lava bowls. 
 sometimes of small size, the lava cones often emit vast volumes of lava, which t 
 the form of broad streams may extend for many miles. The lava pits are by far the 
 
154 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 most numerous group, and most widely distributed throughout the islands. They 
 have no outer slopes whatever, consisting simply of a pit sunk in the ground. Their 
 walls are sometimes vertical, descending without talus to a flat floor ; sometimes the 
 talus is present, and may cover the whole floor, leaving the bottom as a conical pit. 
 Sometimes the walls are inclined, descending at a uniform slope to a flat floor. The 
 slope in this case is usually steep, perhaps 45. The crater rings are the rarest 
 type, and resemble the larger craters found upon the Moon. They have flat floors 
 and sloping inner and outer walls. The crater bowls differ from them in that the 
 bottom instead of presenting a well-defined flattened floor is concave, the curvature 
 being continuous with that of the walls. They are identical in appearance with 
 most of the smaller lunar craters. Section drawings illustrating these different forms 
 will be found on p. 171, and will be described when the various types are reached. 
 Photographs of many of them are also given at the end of this memoir. 
 
 In addition to the craters, there are found numerous other interesting formations, 
 such as lava caves, channels, cracks, spiracles, pinnacles, ridges, etc. A spiracle is 
 literally a blow hole, but in this paper, for lack of a better name, I have used the 
 word to indicate the solid formation surrounding the hole. In dealing with these 
 various objects it has been thought best to describe each class by itself, stating where 
 the best specimens of each may be seen. 
 
 The visitor to Hawaii, on entering the harbor of Honolulu, is at once struck with 
 two very conspicuous volcanic formations, known as Diamond Head and the Punch- 
 bowl. Other smaller and less conspicuous craters, of the same general type, will be 
 found in the immediate vicinity. The Punchbowl, , p. 171, reaches an altitude of 
 498 feet, and is situated within the city limits. The crater is but slightly concave, 
 being filled nearly to the brim, and has a diameter of 2500 feet. The writer did 
 not have an opportunity to examine it carefully, but as it was evidently similar 
 to Diamond Head, 6, p. 171, which was larger, and apparently better preserved, this 
 was not greatly regretted. 
 
 From every direction Diamond Head, Figure 1, presents an appearance similar to 
 a lunar crater. Its highest point reaches an altitude of only 761 feet above the sea, 
 while the diameter of the crater rim measures 3200 by 3700 feet. An ascent of 
 the rim on foot is easily made from a point on the road just beyond the terminus of 
 the electric car line. The rim at this point has an altitude of 450 feet. In the 
 interior of the crater, somewhat to one side of the centre, is located a shallow lake, 
 sometimes dry, whose bed measures 220 feet below the rim where we crossed it. It is 
 surrounded by a very dense growth of thorny shrubs. Within the crater was found 
 
155 
 
 ocean 
 
 PICKERING.- LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 a specimen containing a fossil shell, which was doubtless brought up from the 
 bed by the erupted material when the crater was active. A branching system of 
 cracks, none of them exceeding three inches in width, was found in one place. 
 The inner slopes of the crater range from 20 to 45, the outer from 30 to 7<T 
 Clearly the walls were formerly somewhat higher, and the interior and exterior of the 
 crater about on a level. The edge of the rim is extremely sharp in places. The 
 material is composed of a hardened volcanic mud or tuff, and while the crater some- 
 what resembles numerous of the smaller lunar craterlets, yet their interiors are always 
 at a lower level than the exterior plane on which they are situated, and their inner 
 slopes are steeper than their outer ones. The crater seems to offer little analogy 
 therefore to the formations upon the Moon. 
 
 Cinder cones, c p. 171, form the most numerous class of craters in Hawaii. They 
 are found scattered over the summit of Mauna Kea, in the valley between Mauna Kea 
 and Mauna Loa, in the interior of Haleakala, along the southern and northwestern 
 coasts of Hawaii, and in many other places. A group situated near the summit of 
 Mauna Kea is shown in Figure 2. They have all the characteristics of explosive vol- 
 canoes like Vesuvius, although their craters are larger in proportion to the height of 
 their cones. So far as is known 'they bear no analogy to anything found upon the 
 Moon. 
 
 The third class, or lava craters, on the other hand, present a close resemblance in 
 many respects to some of the lunar formations, and we shall therefore describe them 
 in detail. The first subdivision, the lava cones, are most strikingly represented by 
 Mauna Loa, by far the world's largest volcano. It and Mauna Kea are also our 
 highest mountains if we measure in every case from the mountain's base. For the 
 Hawaiian volcanoes the base lies 15,000 feet below the level of the sea. Nevertheless, 
 the summit crater of Mauna Loa is so large in proportion to its depth that it was 
 thought best to select a small lava cone in Haleakala as the typical example of this 
 form of crater. This cone is shown in the right foreground of Figure 3, and its sec- 
 tion at e on p. 171. The outer slopes of a lava cone are often covered by loose cinders, 
 as in the present case, and the inner slopes may be inclined like those of a cinder cone, 
 although they are generally much steeper, but if the inner walls are of lava its classi- 
 fication is assured. Lava sometimes issues from the summits of these cones, but some- 
 times it comes directly out of the ground, as in Kilauea Iki and at Huehue, no trace 
 of a cone being found. 
 
 Lava has not been known within historic times to overflow the summit crater of 
 Mauna Loa, but it escapes from just below the summit, outside the crater walls, in 
 
156 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 enormous quantities, especially upon the northeastern side. The vast bulk of the 
 mountain seems to have been built up largely from these emissions, and the same is 
 true also of Mt. Etna in Sicily. It is characteristic of these mountains that their 
 slopes are much more gentle than those of cinder cones, and this is especially true 
 of Mauna Loa, Figure 4. The summit of this mountain is 13,675 feet in elevation. 
 The view was taken from the north, and represents the upper 7000 feet of the moun- 
 tain. The summit is very difficult of access on account of the exceedingly rough 
 nature of the ground, the total absence of water, and on account of its flatness of 
 the long distance of the summit from a base of supplies. It is probably best reached 
 from the Kona, or western side, by way of Kealakekua Bay. 
 
 The slopes of Etna are heavily buttressed by ridges, formed each of a separate 
 lava stream, which has flowed from the small lava cones upon the flanks of the 
 mountain. This structure is also well shown in the lunar crater Bullialdus, Figure 5. 
 The diameter of this crater is 38 miles. Since these streams sometimes cross one 
 another, leaving diamond-shaped hollows between them, it is obvious that the 
 formation cannot be due to the grooving of a smooth surface by erosion, but must 
 really be formed by projecting ridges. We shall refer again to this matter in con- 
 nection with Clavius and Kilauea Iki. We thus have indirect evidence of the exist- 
 ence of lava cones upon the Moon, as the source of these streams. 
 
 Until recently this was all the evidence we had. The tall volcanic cone with the 
 comparatively minute crater at its summit, so characteristic of the typical terrestrial 
 volcano, was supposed to be absent from the Moon. In the terrestrial volcano the 
 floor of the crater is always higher than its base ; on the Moon the reverse is true. 
 A recent examination of a lunar photograph taken at the Yerkes Observatory by 
 Professor Ritchey has shown, however, that the terrestrial type of volcano is not 
 wholly absent from the Moon. Craters of this type have not been found before, 
 merely because, like those on the Earth, they are very small. Figure 6 represents 
 the two craters Kies and Mercator. Between them is seen a comparatively small 
 cone with a minute crater upon its summit. It proves to be nine miles in diameter 
 at its base, and 2000 feet in height, while the crater itself measures half a mile in 
 diameter. For purposes of comparison we may say that the diameter of the base 
 of Vesuvius, including Monte Somma, is eight miles, and its height 4000 feet. The 
 diameter of its crater, which varies with every eruption, rarely exceeds one quarter of 
 a mile, and is sometimes but a few hundred feet. The mean angle of the slope of Vesu- 
 vius is 10.7, that of Etna 7.6, of Mauna Loa 5.1, and of the lunar cone 4.8. Vesuvius 
 is partly a lava and partly a cinder cone, which accounts for its steepness. If it were 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 157 
 
 purely a cinder cone its angle might rise to 20 or even 30. Etna and Mauna Loa 
 are both lava cones, the lava of the latter being more fusible. We may infer that the 
 lunar cone is composed of similar material. It is probable that other similar lava 
 cones exist upon the Moon, and one is suspected lying six-tenths way from Copernicus 
 to Kepler, and a little to the north. 
 
 Lunar photographs are usually oriented with south at the top. The right hand 
 side is called east. Figures 14, 15, 31, and 35 were taken with a telescope of long 
 focus, using an aperture of six inches, at the Harvard Station in the Island of 
 Jamaica. The remaining lunar photographs were copied from lantern slides from 
 photographs made by Professor Bitchey at the Yerkes Observatory. In Figure 5, one 
 diameter to the south, and a little to the east of Bullialdus is a pair of coneless lava pits, 
 the southwestern one being much the larger of the two. A few other very minute pits 
 are shown upon the photograph, but all the larger ones have cones. In Figure 29, in 
 the upper left-hand corner, five small craters are shown in a line running north and south. 
 The northern one, which is also the smallest, is coneless. One slightly larger than this, 
 also coneless, is shown just above the centre of the picture, on the southern side of the 
 great rill. It measures five miles in diameter. These seem to be true engulfment cra- 
 ters, as distinguished from the expulsion craters hitherto described. Similar lava pits 
 are found to the west and northwest of Copernicus, and also upon the Oceanus Pro- 
 cellarum. In general they are very minute objects. No large crater pits are known. 
 
 In Figure 7 we have a small terrestrial crater of this type. It is known as 
 Kauhaku, and is found on the island of Molokai. It has no exterior cone whatever, 
 and is merely a hole in the ground. Even explosive craters start in this form, the 
 cone being formed immediately of materials ejected from the hole. Coneless engulf- 
 ment craters abound on the slopes of Hualalai. See k, /, m, and/?, p. 171. Many 
 are also found to the southeast and east of Kilauea, but the best known of all is 
 Halemaumau, on the floor of Kilauea itself, d t p. 171. Figure 8 is known as Kuuohi, 
 or more familiarly as the sixth crater, and is situated six miles to the southeast of 
 Kilauea. See also/, p. 171. The floor of the main crater pit measures about 6,00( 
 feet in length by 2,000 in breadth. Its depth below the surrounding surface is 
 feet. At its eastern end a second crater pit has formed. This measures 2,00( 
 in diameter, and has an additional depth of 600 feet. It furnishes a vertical 
 150 feet in depth of the primary floor, below which the walls form an inverte 
 cated cone to a small floor a few hundred feet in diameter. The lava of the uppe: 
 twenty-five feet of the vertical section has a horizontal stratification, and 
 distinguished from the portion below it. 
 
158 PICKERJNG. LUNAK AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 Of the three great Hawaiian representatives of the engulfment type, Mokuaweoweo, 
 the summit crater of Mauna Loa, Figure 9, and Kilauea, Figure 10, are coneless, while 
 Haleakala, Figure 3, has in places a well-defined outer slope. This might lead us to 
 suspect that it was of composite origin, an impression that is further confirmed by the 
 fact that both its inner and outer slopes are made up in part of lava, and in part 
 of scoria and sand. It seems to have originated partly by the engulfment process and 
 partly by explosions of steam. 
 
 No evidence of the great crater of Mokuaweoweo is to be seen until just before we 
 reach its rim. It measures 3.7 by 1.7 miles in diameter, and is 300 to 400 feet in 
 depth. It is composed of three confluent craters, of which the middle one is much 
 the largest. A portion of the two northern ones is shown in the figure, the view 
 being taken in a direction nearly due south. The crater floor corresponds very 
 closely in its nature to a lunar mare. 
 
 Kilauea is much more accessible than Mokuaweoweo. Twenty-two miles in the 
 train from Hilo, and nine miles by stage brings us to the Volcano House, situated 
 upon its brink. Kilauea consists of a black lava plain measuring two miles by 
 three, bounded on all sides by precipices, often vertical, ranging from 200 to 500 feet 
 in height. The view, Figure 10, was taken from its southeastern rim looking north. 
 It is a curious fact that the black lava usually shows in the photographs to be much 
 lighter colored than the gray cliffs surrounding it. Since the whole lava surface is 
 very irregular, the fact that it is shiny and therefore bright in spots, cannot be given 
 as an explanation of this fact. The surface is distinctly convex, d p. 171, as is the 
 case with the smooth crater floors upon the Moon, and is not unlike in shape and 
 curvature to the flattened summit of Mauna Loa, Figure 4. A in the section 
 indicates the location of the Volcano House. The line B is on a level with it. The 
 highest part of the floor, is near Halemaumau, and in 1905 measured 230 feet above 
 the edge where the Volcano House trail strikes it. The edge is 410 feet below 
 the Volcano House at this point. 
 
 Halemaumau, " the house of eternal fire," is situated three-quarters way across the 
 crater from the Volcano House. It is at present a nearly circular pit, 2,000 feet in 
 diameter, with practically vertical sides, and when the writer was there early in 
 August its depth was estimated at 500 feet. Its floor consisted of a comparatively 
 smooth lava surface, crossed here and there by narrow cracks, which at night showed 
 bright red. That it was liquid only a few inches below the surface was shown by an 
 occasional outbreak, when a flow of lava 5 to 10 feet broad by 20 to 50 feet in length 
 would sluggishly stretch itself across the floor, glow for a few minutes, and then cool 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 159 
 
 and solidify. From a height of 500 feet the phenomenon presented little of interest 
 compared to what had been seen in the last century. The crater is gradually filling 
 up from a subterranean inlet. The depth of the pit in 1902 was estimated at 1,000 
 feet. The lower portion, which has now been filled was then conical in shape. 
 
 Turning now to the third subdivision of the lava craters, the crater rings, we will 
 begin by a study of what is believed to be their internal structure, as exhibited in 
 Figure 11. This photograph represents a vertical section of a small ring crater formed 
 naturally in cooling iron slag. When the slag is drawn off from the furnace it is 
 allowed to solidify in conical moulds four or five feet in diameter, and about a foot in 
 depth at the centre. Unless interfered with, a crater three or four inches in diameter 
 is invariably formed as soon as the surface has fairly hardened. On breaking up the 
 slag considerable cavities are always found beneath the crater. These are well shown 
 in the figure, as are also the cracks connecting them with one another and with the 
 central peak, which it will be noted is also hollow. The large pear-shaped cavity 
 beneath the peak was in the present instance filled up from below with melted iron. 
 It will be noted that the inner walls are very steep, while the outer ones slope more 
 gradually. During the process of formation the crater sometimes fills to the brim 
 and overflows, building up the walls ; later the interior fluid withdraws, forming the 
 crater floor. 
 
 Besides these larger craters other ones are often formed, which, while retaining a 
 base of perhaps two inches in diameter, frequently build up to the height of several 
 inches, forming vertical tubes or spiracles. Sometimes these tubes are closed at the 
 top and sometimes they are left open. For these facts, and for the specimen from 
 which Figure 11 is taken, I am indebted to Mr. J. A. Brashear. 
 
 Halemaumau, known also as " the pit," is the centre of volcanic activity in Kilauea. 
 No eruption has ever been known to overflow the walls of the latter, although lava 
 has sometimes been emitted from cracks located high up on its sides. When the pit 
 of Halemaumau is emptied, it is always through some subterranean passage, occasion- 
 ally reaching the surface, but usually either filling some subterranean cavity, or else 
 discharging beneath the sea. These eruptions, though often accompanied by slight 
 earthquake shocks, have in only one instance caused serious damage and loss of life. 
 In this case it is thought that the active agency was really Mauna Loa, whose erup- 
 tion took place at the same time. 
 
 When Halemaumau is really active the sight is said to be grand beyond description, 
 especially at night. Lakes of liquid lava occur both within and without it. Numer- 
 ous fire fountains from ten to fifty feet in height play over the surface of these lakes. 
 
160 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 At times the surface solidifies, then suddenly a crack will run across it, and in a 
 few minutes the whole solid material will break up into separate cakes which will 
 presently turn on edge and sink beneath the surface of the lake. This again solidifies, 
 and in a few hours the process is repeated. See report of the United States Geo- 
 logical Survey for 1883, p. 106, Major C. E. Button. 
 
 These lakes are especially interesting to the selenographer, since about them are 
 formed crater rings, which seem to be analogous in appearance to the larger crater 
 formations upon the Moon. During the past forty years, since the construction of the 
 hotel upon the rim of Kilauea, they have been very carefully observed, and it will 
 therefore be well in this place to deal with the subject, briefly, from the chronological 
 standpoint. All the earlier descriptions which follow are condensed from the writings 
 of the Rev. Titus Coan. The references given refer to the American Journal of 
 Science. 
 
 On April 2, 1868, a severe earthquake shook the southern coast of Hawaii, and 
 for the next five days a subterranean discharge of lava took place from Kilauea. As 
 a result of this discharge the central area to the northeast of Halemaumau sank about 
 300 feet, carrying with it the vegetation still growing on its surface. The walls of 
 this new pit were inclined from 30 to 60. The lava also flowed out of Halemaumau, 
 leaving a circular pit 3000 feet in diameter at the top, 1500 feet at the bottom, and 
 500 feet deep. Its walls were in some places vertical, and in some slightly over- 
 hanging. A. J. S., XCVII, 96. In A. J. S., CXVIII, 227, it is stated that the 
 depth was 400 feet and that the distance across the bottom was one mile less one 
 hundred feet. It is also stated that the pit had formerly been filled up not only from 
 the bottom, but by lateral discharges from the walls. 
 
 A year later Halemaumau had filled to within 100 feet of the top, the level area 
 within it showing eight small apertures within which the liquid lava could be seen 
 boiling fiercely 50 to 100 feet below the surface. A few months later the lava was 
 within 25 feet of the rim, and the diameter of the pit was said to have enlarged to 
 over a mile. A. J. S., XCIX, 393. 
 
 In 1870 the pit overflowed, the lava pouring down and partly filling the north- 
 eastern depression. At the time of an eruption such as this, the lava rises, overflows 
 and cools, thus forming a raised rim or circular dam. Such a rim is shown on a 
 small scale in the slag crater, Figure 11, and on a much larger scale in the photo- 
 graphs of Halemaumau, Figures 12 and 13, the cakes of lava there represented 
 appearing much like broken cakes of ice. In Figure 14 is shown a portion of the 
 Moon near the limb, so as to present the craters obliquely. It will be noted that the 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 161 
 
 two large craters there represented, named Schickard and Phocylides, both present a 
 form similar to the crater rings of Halemaumau. The chief one, Schickard, measures 
 134 miles in diameter. All of the larger craters on the Moon are of this type. 
 When seen towards the centre of the disc, however, their depth appears by an optical 
 illusion greatly exaggerated. Thus Clavius, the largest crater shown in Figure 16 
 measures 143 miles in diameter, and judging by appearances only might be 15 to 20 
 miles in depth. Its depth actually measures two and a half miles, or about the same 
 as the diameter of one of the most minute craterlets visible in its interior. The slope 
 of its eastern inner wall is about 10. 
 
 There is another crater known as Wargentin, lying between Schickard and - 
 Phocylides and the limb, Figure 14. It is not shown in the photograph because it is 
 a very difficult object. This is not on account of its size, since it measures fifty-four 
 miles in diameter, but because it has no interior depression. In this respect it has 
 heretofore been thought to be unique upon the Moon. That such is not the case, 
 however, an inspection of Figure 32 will show. Near the upper edge of the figure 
 a large low nameless crater is to be seen whose interior is obviously at a greater 
 elevation than its exterior, although the difference is not as great as in the case of 
 Wargentin. It measures about 'fifteen miles in diameter. In each of these cases 
 the lava passage leading to the interior of the crater evidently became choked while 
 the crater was still brimful of molten matter, thus preventing the withdrawal of the 
 lava, and preserving the crater as a permanent illustration of the method by which 
 these formations are produced. 
 
 The outside height of the crater rings in Halemaumau rarely exceeds 15 to 25 feet. 
 The inside height is constantly varying with the fluctuating level of the surface 
 of the lava lake. When the outside height becomes too great to withstand the 
 internal pressure, the rim gives way, and the lava breaks through and floods the 
 surrounding regions. By means of this successive building up and flooding, the whole 
 region around Halemaumau was elevated, until the walls became too high and too 
 thick for the floods to escape over or through them. The pit at this time measured a 
 mile by a mile and a half, and was 700 feet in depth. Since the lava could not 
 now overflow the rim, it escaped by subterranean passages, and thus flooded the 
 other portions of the floor of Kilauea. A. J. S., CII, 454. 
 
 It is doubtful if the walls of the pit were raised exclusively by the process thus 
 explained by Mr. Coan. It is probable that the whole of this portion of the floor 
 was also elevated as one piece by the pressure of the subterranean lava, as occurred 
 
 in the later eruptions. 
 
 11 
 
162 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 On the Moon the height of the outer wall is roughly proportional to the diameter 
 of the crater. For a crater whose diameter measures 30 to 40 miles, the height 
 of the outer wall is usually about one mile. For a crater a mile in diameter its 
 height would be 150 feet. Dividing this figure by 6, the correction for gravity 
 mentioned at the beginning of this paper, we find 25 feet to be the theoretical 
 height for a terrestrial crater a mile in diameter, thus agreeing with the figure 
 above given. 
 
 In April, 1879 a silent discharge from the crater occurred, the lava apparently 
 making its escape out at sea. A few months later activity was again resumed, the 
 crater becoming extremely active in 1880. A. J. S., CXVIII, 227 ; CXX, 72. 
 
 On March 6, 1886, Halemaumau was again emptied. It was a silent discharge 
 like its predecessor. For several days the surrounding walls continued to fall into 
 the pit. A month later the deepest portion had the shape of an inverted cone, 
 whose apex was 570 feet below the floor of Kilauea. Three months after an erect 
 cone was found formed of loose blocks, and measuring 150 feet in height. During 
 the next two years this cone gradually floated upwards, no additions being made 
 to its summit. This action seems to have been due to the pressure of the lava 
 beneath it. The rate of elevation was about three inches per day. A. J. S., CXXXI, 
 397; CXXXVII, 48. 
 
 The next discharge of Halemaumau occurred on March 6, 1891. It was accom- 
 panied and followed by a series of light earthquake shocks, but otherwise it was a 
 quiet discharge. Prior to this date the pit contained a central cone with three lakes 
 surrounding it, one east, one west, and one south. The direction of the current on 
 the surface of these lakes in each case was away from the cone. The cone which con- 
 sisted in part of several peaks rose to a height of at least 200 feet above the lakes. 
 As'seen from the Volcano House about one-third of its height was above the western wall 
 of Kilauea. The structure of these peaks was loose, and sulphurous vapors escaped 
 from their whole surface. The fire fountains on the lakes sometimes played obliquely, 
 and without the emission of steam. On the evening of March 6, at 9:30, a light earth- 
 quake shock was felt, and the peaks settled slightly ; the next morning they were out 
 of sight. A month later in place of the peaks and lakes an empty pit was seen, 
 measuring half a mile in diameter. The walls were vertical and 500 feet in depth. 
 Soon after this the lava reappeared in the bottom of the pit, and by the end of April 
 a lava lake 100 to 200 feet in diameter had formed. A year later the diameter of 
 the lake was a little over 800 feet. It was then very active, as many as fifteen fire 
 fountains having been counted at one time. It was probably at about this time that 
 
PICKERING.- 
 
 LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 163 
 
 the photographs shown in Figures 12 and 13 were taken. A. J. S., CXLI 336 507 
 CXLII,77; CXLV,241. 
 
 The next discharge was on July 11, 1894. In August, 1892, the edge of the pit 
 Halemaumau was 282 feet below the level of the Volcano House. The surface of 
 
 the lake was 240 feet below the edge, 522 feet 
 in all. (See cut.) In March, 1894, the surface 
 of the lake was 75 feet below the Volcano 
 House, making a total rise of 447 feet in nine- 
 teen months, or about ten inches per day. In 
 1892 the lake was in the bottom of the pit. 
 
 In 1894 the pit was filled, and the lake was on the top of a flat hill covering the pit 
 and situated 207 feet above its former rim. The pit was filled partly by overflows from 
 the lake, and partly by a rise of the whole bottom of the pit. The lake now measured 
 800 by 1200 feet. 
 
 On March 21 an area measuring 400 by 800 feet, situated on the northern bank of 
 the lake, was suddenly elevated eighty feet above the other banks. The raised area 
 was much shattered. It subsequently sank gradually, until on July 11 it again 
 reached its former level. On that date the lava in the lake began rapidly to sink, 
 and the walls about the lake to crack off and fall into it. The lava sank at the rate 
 of twenty feet an hour until eight that evening. From noon till eight there was 
 scarcely a moment when the crash of the falling blocks was not heard. A number of 
 times a section 200 to 500 feet long, 100 to 200 feet high, and 20 to 30 feet thick 
 would drop with a tremendous roar into the boiling lava. Such a section would form 
 for a time a floating island in the lake, but would subsequently dissolve and sink. 
 The grandeur and magnificence of the scene at night were indescribable. Meanwhile 
 the fountains in the lake continued to play as if nothing unusual were happening. Only 
 a few slight earthquakes accompanied this discharge. A. J. S., CXLVIII, 338. 
 
 Since this time for an interval of over twelve years the volcanic forces in Kilauea have 
 been practically quiescent. There was a slight display of activity in 1896, and again in 
 1897, but the pit remained empty, and no activity whatever has been seen since then, 
 except for the gradual and uneventful filling of the pit which seems now to be taking 
 place. No such protracted interval of quiet has been known heretofore, the longest pre- 
 vious period amounting to only a few months. It is of interest to note that of the five 
 recorded discharges, four occurred during the rainy months of March and April. 
 
 From these descriptions we can obtain an idea of how the lunar craters were 
 formed, and can also account for the flat-topped vertical cliffs that we find about 
 
164 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 some of the maria, as in the case of Sinus Iridum, Figure 15. The walls of Mokuaweo- 
 weo or of Kilauea, Figures 9 and 10, furnish excellent illustrations of these forma- 
 tions. Examples of the crater rings themselves, however, are rare. The best preserved 
 one that we were able to visit was found on Hualalai, at an elevation of 400 feet 
 below the summit of the mountain, and was reached a few minutes after passing the 
 so-called Bottomless Pit. It was located on the floor of a crater some 500 feet in 
 diameter by 100 feet in depth, g p. 171. The diameter of the crater ring was 120 
 feet, its internal depth was six feet and the height of its outside wall twelve to sixteen 
 feet. It is shown upon a larger scale at h p. 171. Near the centre of the outer crater, 
 and outside of the crater ring, was a low peak. What appeared to be another crater 
 ring was seen from the distance of a mile on the northern slopes of Mauna Loa, and 
 will be described later. A portion of one of the crater rings formed by Halemaumau 
 is still to be seen on the rim near the view-point where visitors look down into the 
 interior. 
 
 The reason that these crater rings are so numerous upon the Moon and so rare 
 upon the Earth is apparently that the terrestrial ones are not generally permanent. 
 The smaller craters on the Moon do not take this form, and if some of them did exist 
 formerly they have since been destroyed. The reason of this is that when the lava 
 recedes into the bottom of the pit, the depth is so great in proportion to the diameter, 
 that the walls cave in, destroying the ring, as usually happens at Halemaumau. 
 On the Moon no crater is known whose depth exceeds five miles, and two miles is the 
 usual depth for large craters. This distance compared to a diameter of twenty to 
 sixty miles is, so slight that the ring remains uninjured. The outer walls of Haleakala 
 may in part be the remains of an old crater ring of very elliptical shape. They have 
 been breached and totally destroyed in two places. 
 
 The floors of the craters on the Moon are of three kinds, either they are furnished 
 with a central peak, like Tycho, the large crater in the lower left hand corner of 
 Figure 16, or they contain one or more smaller craters, in general not central, like 
 Clavins, in the same figure ; or they are without conspicuous detail, like Kies, Figure 
 6. In the last two cases the floor is often of a later origin than the walls, as indicated 
 by its color and smoothness, the original floor having been melted by a flood of dark 
 colored lava from below, which dissolved all the lowermost portions of the solid crust 
 with which it carne in contact. This seems to be the case with Longomontanus, the 
 large crater northeast of Clavius. Its diameter is ninety-one miles. Often, however, 
 the floor is bright, and not perfectly smooth, as in Tycho, showing it to be part of the 
 original formation. Central peaks are occasionally, but by no means universally, 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 165 
 
 found upon these original floors. Indeed they rarely occur in craters of less than 
 Four miles in diameter. Longomontanus and another crater known as Pitatus present 
 the very unusual phenomena of eccentric internal peaks. Equally striking is the fact 
 that in each case the peak is found in connection with a dark floor. The explanation 
 in both cases is probably that the lava, after dissolving the original floor, had begun 
 to dissolve the peaks, which were pushed by the lava currents to one side, where 
 cooling and solidification set in before the process was completed. In both these 
 cases the peaks are unusually small in proportion to the size of the craters, as would 
 naturally be the case if they had been floating partially submerged in the lava. 
 
 Central peaks are seldom found in the Hawaiian craters, probably because the 
 latter are so small. The best illustration seen was that of the small crater already 
 mentioned at the northern base of Mauna Loa, about eight miles west of the Humuula 
 sheep ranch, i p. 171. Unfortunately we could not get within a mile of it, but it 
 seemed to be well defined. Its diameter was thought to be about 450 feet. The 
 walls were red, and the central peak dark brown. The height of the peak was the 
 same as that of the walls. Another small crater two or three hundred yards back 
 of Mr. Maguire's at Huehue in Kona showed a smooth central peak 15 feet in height. 
 It was completely grass grown. 'One or two of the small craters on Hualalai, 
 k p. 171, showed the same formation. The peaks were never pointed, but sandy and 
 rounded as in the slag crater shown in Figure 11. A crater containing a central 
 peak or craterlet is said 'to exist half a mile beyond the sixth crater near Kilauea. 
 Near the centre of Haleakala is found a straight, narrow ridge, 150 feet in height by 
 400 feet in length, along its crest. Its sides slope at an angle of about 30 and it is 
 composed apparently of gravel and scoria. At its eastern base is the cave where 
 parties sometimes pass the night. Although not especially conspicuous among the 
 crater cones which dot the floor, some of which are 500 feet in height (see Figure 3), 
 it seems to be unique in shape, running lengthwise of the crater. It is also almost 
 exactly central in position. In passing it may be stated that the maps of Haleakala 
 give a very erroneous impression of the shape of its floor. The floor at the Koolau 
 and Kaupo gaps falls off sharply, showing the outline of the true floor to be not 
 S-shaped, but elliptical, and extending nearly due east and west. It is an ellipse 
 of great eccentricity, the length of the floor being about four times its breadth. 
 
 Figure 18 represents a portion of the middle crater of Mokuaweoweo. Somewhat 
 nearer than the centre is shown an active cinder cone composed apparently of a 
 medium-sized crater and two or three smaller ones upon its rim. We were not able 
 to visit it. Like the craterlets found in Haleakala, it reminds us of those found in 
 
166 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 Clavius. The photograph was taken in 1903 by Mr. C. W. Baldwin, and was 
 forwarded to me through the kindness of Professor W. D. Alexander. 
 
 One of the most interesting craters that we visited was Kilauea Iki, or little 
 Kilauea. It is situated about a mile from the Volcano House. The floor is level, 
 one-quarter of a mile in diameter and 750 feet below the rim. The walls are very 
 steep, but can be descended in certain places with care. Numerous small craterlets 
 are scattered irregularly over the floor. The most complete of these is shown in 
 Figure 19. Its height was 15 feet, and the diameter of the rim, which was composed 
 of lava of a somewhat ropy appearance, 25 feet. A stream of lava had poured from 
 the summit, but did not get far beyond the rim. There may have been as many as 
 fifty rudimentary craterlets scattered over the floor, in all stages of growth, from 
 a hardly noticeable elevation to the complete craterlet shown in the figure. 
 
 The process of construction was clearly shown, and was probably identical with 
 that which produced Halemaumau, and Kilauea itself in the first place. In Figure 20 
 the top of one of the other craterlets is represented in the foreground. It was taken 
 near at hand from the summit of the first one, and its development is clearly incom- 
 plete. The surface of the crater floor of Kilauea Iki seems to have solidified into 
 a layer six to ten inches in depth and distinct from the portions below it, much as in 
 the case of the sixth crater, Figure 8. A liquid core forced up from below raised 
 this surface layer locally, and shattered it into separate pieces like cakes of ice. 
 This core in the case of some of the smaller craterlets was sometimes only two or 
 three feet in diameter, and could be seen beneath the shattered surface. In one 
 instance its summit seemed to have an almost globular form, five feet in diameter. 
 
 If the volcanic forces beneath these craterlets had been more intense, it is probable 
 that the issuing lava would have completely destroyed them, forming a series of 
 crater pits, into which the lava would have subsequently retreated. In the south- 
 eastern part of the floor two such pits were found, perhaps 15 feet in depth by 
 30 in diameter, down into which a stream of lava had poured, but had solidified 
 without filling them up. One of these pits is shown in Figure 21. Figures 20 
 and 19 therefore illustrate the earliest stages of formation of a lunar crater. No 
 other example in Hawaii is known to the writer which shows as satisfactorily as this 
 the irregular distribution of craterlets over a crater floor. 
 
 A low ridge due to compression, caused by the sinking of the convex surface 
 crosses the eastern end of the crater floor. Similar ridges are seen on some of the lunar 
 maria, notably on Serenitatis, Figure 17. Two short, clearly marked ridges project 
 onto the southern side of the floor of Kilauea Iki, caused by lava streams which had 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 167 
 
 descended from the crater wall. They resemble similarly located formations found 
 in the lunar crater Plato. In Clavius, Figure 16, similar ridges are seen projecting 
 from the outer slopes of the two chief craterlets upon the northern and southern 
 walls. Similar ridges, although much more complicated in structure ou account 
 of their numbers, have already been described in Bullialdus, Figure 5. 
 
 Leaving the craterlets and ridges of Kilauea Iki, and proceeding along a defile 
 towards Kilauea itself, three successive lava dams were reached, each of which had 
 served to hold back a small lava lake. In construction they were similar to the 
 circular one represented in Figure 12, except that they were straight, and merely 
 stretched across the defile. The first lake measured 400 feet in length by 150 in 
 breadth. The second dam rose eight feet above its surface, and three feet above 
 that of the second lake. By the side of this lake a core of lava of the most brilliant 
 colors red, yellow, brown, and purple had escaped from the ground, and from it 
 a black lava stream had descended to the surface of Kilauea Iki, 200 feet below. 
 A similar core but without the colors was seen at Huehue. Both these flows occurred 
 during the last century. Both were small, and in neither case did a cone appear, 
 the lava issuing directly from the ground. 
 
 The fourth subdivision of the /lava craters, described earlier as crater bowls, is 
 illustrated in Hawaii by what is known as Aloi, or the Third Crater, near Kilauea. 
 Other illustrations are found on the slopes of Hualalai. A crater near the summit, 
 j, p. 171, was estimated at 800 feet in diameter by 200 in depth. The sloping walls 
 were of lava, and the bottom of sand. The comparatively shelving outer walls 
 probably did not exceed 100 feet in height. A portion of the interior of this crater 
 is shown in Figure 22. A somewhat larger crater bowl with much steeper walls 
 is found on the summit. With favorable definition such a crater would be readily 
 seen upon the Moon, and could not be distinguished in any way from many others 
 found there. 
 
 The largest craters on Hualalai occurred near the summit, and shortly before 
 reaching the top we crossed a lava field strongly resembling a small lunar mart. 
 A far larger number of craters are found on this mountain than on any of the 
 others, more perhaps than on all the others combined. The three types, of cinder 
 cones, pits, and bowls, are each represented by numerous examples. One of the 
 craters that we passed after leaving the summit, k, p. 171, had sloping walls and a flat 
 floor with sand hills on the bottom. One of these was twenty feet in height. Near 
 the base of the walls was an inner terrace extending all around the crater. This 
 feature of a single, well marked, inner terrace is conspicuous in a considerable 
 
168 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 number of the lunar craters, such as Fabricius, Hercules, Manilius, Reinhold, and 
 Bullialdus (see Figure 5). 
 
 Some of the lava pits occur very near together, the intermediate wall being only 
 a few yards in thickness, but we saw none which actually intersected. When the 
 large summit crater was formed a smaller one near it was partially destroyed, and 
 filled by the erupted material from the larger crater. 
 
 We have no description of the method of formation of a crater bowl, but a study 
 of the various sections on p. 171 will show how they have probably been formed, 
 Starting with a crater ring such as is shown in Figure 19, and in section at h, p. 171, 
 when the lava retreats a crater pit will be formed within it. If the pit is very deep 
 relatively to the diameter of the ring, the latter will be destroyed, and we shall 
 have simply an ordinary pit, I. If the pit is not so deep relatively to the diameter 
 of the rim the latter may be preserved, n. The floor of the pit may be flat, or it may 
 sink towards the centre, m. If the pit is on a large scale the sides are less liable 
 to be vertical, and moreover a talus will collect at their base, o. This will 
 gradually become rounded as at p, or if the crater ring is still left, as at /. The 
 flattened floors of the larger craters on the Moon are illustrated by #, while the 
 terrace and central peak are shown at k. 
 
 With the exception of the crater pits, nearly all the smaller depressions upon the 
 Moon are crater bowls, and they outnumber at least ten times all the other depres- 
 sions put together. The smallest crater rings are about five miles in diameter. 
 One of the largest and best situated crater bowls is Triesnecker, 14 miles in diameter. 
 It has an inconspicuous central peak. In the smaller bowls this feature seems to be 
 lacking. A well-graduated series of bowls is shown in the interior of Clavius, 
 Figure 16. The process which converts a pit I into a bowl p upon the Earth is due 
 chiefly to the action of water. Upon the Moon, even in former times, water was 
 probably scarce, but owing to the extremely rarefied atmosphere the extremes of 
 temperature are excessive. A range of 300 C. or 540 F. occurs every fortnight, 
 and it seems likely that a considerable destruction of ridges and filling of hollows 
 would be due to this cause by itself. 
 
 Another explanation of crater bowls is given by Gilbert in his dissertation on 
 " The Moon's Face," Philosophical Society of Washington, Bulletin XII, 251, where 
 he suggests that they may be due to a single explosion of steam, like the terrestrial 
 " maars." This seems improbable, since volcanic features due to steam are notably 
 absent from the Moon. On Hualalai the crater bowls have smooth lava walls which 
 are not at all fragmentary, as would be the case were they due to an explosion. 
 
1'ICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 169 
 
 In drawing the sections on p. 171, it was necessary to represent them on several 
 different scales. The smallest scale adopted was T J TT or a quarter of an inch to 
 1000 feet. The dimensions of all save a, b, and d are based only on estimates, but 
 these were made on the spot with all possible care, excepting c and e, where the 
 estimates were based on photographs, and o, for which I had to depend on my 
 memory. All the sketches were either made on the spot or were taken from 
 photographs. The craters on Hualalai are designated for lack of a better system of 
 nomenclature in the order in which we visited them from Huehue. Number 8 is 
 the summit crater. In the description which follows, the length of 1000 feet is 
 given in each case in inches as measured on the section. 
 
 a Tuff cone. Punch Bowl, inch. 
 b Tuff cone. Diamond Head, \ inch, 
 c Cinder cone on Mauna Kea, 1 inch. 
 d Lava pit. Kilauea, \ inch. 
 e Lava cone in Haleakala, 1 inch. 
 
 / Lava pit Kauohi or Sixth Crater near Kilauea, \ inch. 
 g Lava cone and ring. Crater number 10 on Hualalai, 2 inches. 
 h Lava ring. On floor of crater number 10 on Hualalai, 8 inches. 
 i Lava ring on Mauna Loa, 2 inches. 
 j Lava bowl. Crater number 9 on Hualalai, 1 inch. 
 k Lava pit. Crater number 11 on Hualalai, 2 inches. 
 I Lava pit. Crater number 3 on Hualalai, 2 inches. 
 m Lava pit. Crater number 6 on Hualalai, 2 inches. 
 n Lava pit. Crater number 12 on Hualalai, 1 inch. 
 o Lava pit. Alealea or Fourth Crater near Kilauea, inch. 
 p Lava pit Crater number 7 on Hualalai, 4 inches. 
 
 After the craters, among the most important features of the lava flows are the 
 elevated formations, the spiracles, pinnacles, and ridges. When the gases work 
 their way up to the surface from a subterranean cavity they escape by little 
 apertures called blow holes. In so doing they often carry small quantities of 
 lava along with them. This lava quickly hardens on reaching the surface and 
 builds up a tube around the aperture which we have called a spiracle. Some- 
 times it is closed at the top by the last escaping lava, and sometimes it is left 
 open. These spiracles are found of all sizes, from one measuring three or four 
 inches in diameter, up to another measuring one hundred feet. The former, found 
 in Kilauea, was twelve inches in height, and contained a hole one inch in diameter 
 running its whole length, except where it was closed at the top. The latter was found 
 on Hualalai, and will be described presently. 
 
170 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 Figure 23 represents a spiracle fourteen feet in height by six feet in diameter at 
 its base, found in Kilauea near Halernaumau. It is built up of what may be described 
 as great drops of solidified lava. The interior tube is open at the top and measures 
 a foot in diameter. Another somewhat smaller spiracle is seen in the background. 
 Often several spiracles occur side by side with confluent bases, as in Figure 24. This 
 object is also located in Kilauea, near the corral where the horses are left. It closely 
 resembles the central j>eak or range of peaks so often found in the lunar craters, and 
 is doubtless due to the same cause. See Tycho and Longomontanus, Figure 16. It 
 is ten feet in height. Several of the spiracles forming it are open, and several are 
 closed. There are a number of large cavities in the interior, in some of which were 
 found some very slender lava stalactites. No lava flow had escaped from any of the 
 craters, but two outbursts had occurred upon the side, and may be seen about half 
 way down the slope, below the right-hand summit of the ridge. 
 
 Figure 25 shows a much larger row of spiracles found on Hualalai. They measure 
 about a thousand feet in height above their base. Midway between the two highest 
 summits are two smaller ones. The left hand of these is known as the Bottomless 
 Pit. The little cone measures one hundred feet in diameter at its base by sixty feet 
 in height. A narrow tube a few yards in diameter opens at the summit, and it is 
 said that it has been sounded for 1400 feet without reaching bottom. Whether 
 this figure is correct or not, doubtless the tube is very deep, and no bottom is 
 visible. These spiracles equal in height many of the central peaks found upon the 
 Moon. 
 
 Sometimes a row of small conical elevations, about equally spaced, occurs upon 
 the Moon. Such a row is found in the eastern part of the floor of Wilhelm I, the large 
 crater shown in the lower right-hand corner of Figure 16. The illustration is on too 
 small a scale to show them to advantage, however. A row of still larger cones is found 
 just outside and northeast of the crater. They seem like spiracles thrown up along 
 the course of a steam crack. 
 
 Sometimes the lava slabs pile up on one another in horizontal layers, as in 
 Figure 26, and sometimes much more irregular blocks occur, without any apparent 
 order. These form pinnacles with very steep sides and ends. Their origin seems to 
 be due to recent flows of lava which have transported and piled up the fragments 
 formed from the earlier flows, somewhat as the ice pack is transported by the winds 
 and currents in the far north. This object was found in Kilauea. 
 
 Another type of pinnacle consists of a single block of lava which may rise as high 
 as sixty feet above the surrounding plain. The sides are often precipitous, and there 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 171 
 
 n 
 
 T 
 
 SECTIONS OP CRATERS. 
 
172 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 is no summit crater. It is probably a solid block fallen from a neighboring cliff, that 
 had been undermined by the liquid flow, and after floating awhile and being trans- 
 ported, was now frozen in, in its present position. Such a block is shown in Figure 
 27. It is twenty-five feet in height, and was found upon the floor of Haleakala. A 
 second one is shown in the distance. Similar objects are of frequent occurrence upon 
 the various maria of the Moon. Doubtless they are often formed as above described, 
 but in many instances it is evident that they have been left in their original positions, 
 while the objects formerly surrounding them have been destroyed by the flood of 
 molten lava. Innumerable pinnacles are found upon the Mare Imbrium. A num- 
 ber of them are shown in Figure 28. The large crater in the photograph, near the 
 left-hand edge, is Euler. Its diameter is nineteen miles. 
 
 A curious feature in Hawaii is the very extensive series of caves that penetrate 
 the lava, especially the flows from Manna Loa. Indeed, so many of them have been 
 found that it has been suggested that they make up an appreciable part of the bulk 
 of the mountain. A very accessible cave is situated a few miles above the town of 
 Hilo. It is said to extend two miles up and two miles down the mountain from the 
 entrance, which is a place where the roof happened to fall in, disclosing the cavity. 
 The breadth of the cave is about thirty feet. Its height varies in the portion that 
 we traversed from three to ten feet. Larger caves are found in other places, some- 
 times, according to Button, being as much as sixty or eighty feet in height, and wide 
 in proportion. Their origin is due to the fact that the surface of the lava hardens 
 first, and that the lower portions meanwhile flow away, leaving the cavity. Small 
 caves occur on the floors of Kilauea and Haleakala where, since the floors are level, 
 the formation seems to be duetto the collecting of gases under sufficient pressure to 
 hold back the lava until it has had time to solidify. Sometimes lava channels form 
 without any roof. Some well marked channels and caves are found two or three 
 miles north of Huehue on the Kona coast. A lava channel was noted not far from 
 the summit of Hualalai, where the path crosses an open lava field. Another channel 
 was found in Kilauea near Halemaumau. 
 
 At first it was thought that these lava channels were analogous to the broad 
 grooves found upon the Moon, of which the valleys of the Alps and of Rheita are 
 the most conspicuous examples. The Valley of Rheita is shown in Figure 31. It is 
 190 miles long by 15 miles wide. Several parallel valleys similar to these are found 
 to the southwest of Pallas, and a less well marked series to the southeast of Sinus 
 Iridum. The great range of the Altai Mountains in the southwestern quadrant of 
 the Moon seems to form one side of such a valley constructed upon a very large 
 
PICKERING.- LUNAR AND HAWAIIAN PDYSICAL FEATURES COMPARED. 1 73 
 
 scale. If so, the other side must have been destroyed by a subsequent melting 
 leaving an unusually smooth, light-colored surface in its place. 
 
 It was later concluded that these valleys were produced by a continuous faulting 
 ong a line of volcanic weakness, and were therefore analogous to the craters where 
 instead the faulting extends in all directions from a volcanic centre. The best 
 Hawaiian representative of these grooved valleys is therefore probably the great 
 crater of Haleakala (Figure 3), whose length measures seven miles and its breadth 
 These are about the relative proportions found in many of the valleys south- 
 west of Pallas, nor are their dimensions so very diflferent. The line of six small 
 craters found in the bottom of Haleakala corresponds to the similar line of small 
 craters found along the minute rill in the bottom of the Valley of the Alps. 
 
 The fact that often one and sometimes both ends of the lunar valleys are closed 
 by high walls, as is the case with Haleakala, strengthens the second explanation of 
 their origin as opposed to the earlier one. Elongated craters forming an intermediate 
 step between the ordinary craters and the grooved valleys are of frequent occurrence 
 upon the Moon. The largest of these is Schiller, in the southeastern quadrant. A 
 nameless one is shown in the lower left-hand corner of Figure 31, and another in the 
 lower right-hand corner of Figure 29. Others are shown on the border of the mare 
 in the same figure. 
 
 The lunar rills may be divided into two classes, rills and crater rills. The rills 
 proper are extremely numerous upon the Moon. About a thousand are already 
 known. The Ariadaeus rill, shown in Figure 29, is the widest and most conspicuous 
 of them. It measures three miles in breadth by a little over half a mile in depth, as 
 determined by the shadows of the ridges that cross it in various places. Like all true 
 rills its course is approximately straight, or made up of curves of long radius. In its 
 bottom are several minute craterlets not shown in the photograph. Evidently like 
 our dikes and mineral veins it has been partially filled from below. Other narrower 
 rills, apparently bottomless, are found on the Moon. Two much smaller parallel rills 
 with a north and south direction are found upon the mare to the left. One of these 
 is faintly shown in the photograph. The general view that the rills are simply 
 cracks in the lunar surface is undoubtedly correct. They occur most frequently 
 in formations of the secondary period, that is in the dark surfaces, or if found in the 
 primary formations, it is where the surface has apparently been softened and par- 
 tially flattened out by the application of heat, as in the present instance. Rills 
 are frequently found at the edges of the nutria and running parallel to them, as in 
 Serenitatis and Humorum. 
 
174 PICKEKING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 A large crack is found in Kilauea in precisely this position, Figure 30. It is from 
 6 to 8 feet wide and from 20 to 30 feet deep near the bridge. It is said to be 
 about a mile in length. A crack 5 to 20 feet in breadth, and 40 to 200 in depth, 
 by 16 miles in length is located southwest of the crater, and a similar one 
 parallel to it is found near by. Several cinder cones occur upon these cracks much 
 as crater pits do upon the Moon. The cracks themselves have been partly filled up, 
 but one said to be 1500 feet in depth and 5 to 15 in width is situated not far from 
 the Sixth Crater near Kilauea. 
 
 Keanakakoi is a small crater one and a half miles southwest of Kilauea Iki. 
 Its floor measures 500 feet in diameter and is 300 feet below the rim. It illustrates 
 the craters having smooth floors upon the Moon. The lava surface itself is wonder- 
 fully smooth, but a close inspection shows that it has a convex surface, rising from 
 10 to 15 feet higher at the centre than at the edges. In this respect it also 
 resembles what we find upon the Moon. The surface is slightly undulating, the 
 hillocks measuring perhaps two feet in height above the depressions, thus indicating 
 compression of the floor. The surface is also everywhere seamed with cracks, from 
 one to three inches in breadth, and running in all directions. This indicates subse- 
 quent contraction. As often occurs upon the Moon, a crack was found running 
 parallel to the edge of the floor, and not far from the walls. Seven prominent radial 
 cracks were counted. The arrangement strikingly resembled that of the rills in 
 Gassendi. A good map of this crater is given in "The Moon," by Neison, p. 337. In 
 no place did the cracks exceed eight inches in breadth. Ferns are beginning to 
 grow in these here and there. 
 
 A very different type of crack is sometimes produced where the surface is forced 
 open by a subterranean lava flow, and small craters and blow holes are formed 
 along its length. Such a one is found at Huehue, due to an eruption on the slopes 
 of Hualalai in 1801, Figure 33. A ridge 50 feet in height in some places, and 
 carrying on its summit a crack 30 feet wide by 40 feet deep, and extending for 
 perhaps a mile has been produced. It is very irregular in outline. 
 
 Such a crack is represented upon the Moon in Figure 32. It is a good illustration 
 of the crater rills so called, and is known as Bullialdus <. The straight portion 
 measures 40 miles in length by perhaps one and a half in breadth. Its edges are 
 elevated as in its Hawaiian representative, and are quite as irregular in outline in 
 proportion to its size. The distinction between these two types of rills is not sharply 
 drawn upon the Moon, and the same rill sometimes exhibits both types in different 
 portions of its length, as is the case with three small rills located on the northeastern 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 175 
 
 and southeastern flanks of Copernicus, within one diameter of the crater rim. A 
 much larger and more conspicuous crater rill occurs one and a half diameters to the 
 northwest of Copernicus. The craterlets in this case are so distinct, however that 
 the rill-like character is not so well marked. 
 
 Another type of depression that is found upon the Moon is known as a river-bed 
 from its resemblance to its terrestrial analogue. Thirty-four of them 
 have been catalogued. They have been so fully described else- 
 where, Annals of the Harvard College Observatory, XXXII, 84, that 
 it is unnecessary to more than refer to them here. The figure 
 represents one of the larger ones. It is found on the slopes of Mt 
 Hadley. Its length in a straight line is 50 miles, and its maximum 
 breadth 2000 feet. It tapers uniformly from one end to the other. 
 In Figure 6 a marking is found closely resembling one of these river- 
 beds. It is situated due south of Kies and west of Mercator. Its 
 true character can only be ascertained by visual observations made under excep- 
 tionally favorable atmospheric conditions. 
 
 A favorite argument of those who deny that water ever existed upon the 
 Moon is the statement that if such were the case, signs of erosion would be found 
 upon its surface. In the case of the Earth, where vast bodies of water are present, 
 these signs are very pronounced in the eroded valleys of mountain regions, and the 
 alluvial plains of the more open country. When we search the coarser detail upon 
 the Moon no such signs are to be found. This makes it certain that large quantities 
 of water could never have been found upon its surface, nor indeed should we 
 expect such to be the case considering the small value of the force of gravitation 
 existing there. If the Moon ever possessed any water at all, it must have been in 
 comparatively small quantities, and we should accordingly look among its finer detail 
 for any evidence of its former existence. 
 
 Figure 34 represents Theophilus, a crater some 64 miles in diameter. The 
 central peaks rise 5000 to 6000 feet above the crater floor, and are indented by 
 numerous deep valleys, four being clearly shown in the photograph. It is believed 
 that these valleys are due to erosion, and are analogous to those shown in Figure I 
 This figure represents a mountain ridge just back of Honolulu, and was taken from 
 another ridge called Tantalus. Similar valleys occur on the central peak of Erato 
 thenes In the case of Copernicus they have cut so deeply that they have acti 
 divided the central mass into three distinct mountains. The precipitation canno 
 have come from a general atmospheric circulation, but more likely from steam 
 
176 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 expelled from the openings of the spiracles in the central peaks themselves. The 
 valleys seem to be flat bottomed, which would imply the action of ice rather than 
 water. Indeed, at the present time the central peaks of Theophilus are of a dazzling 
 brilliancy as compared with their surroundings. This, it is believed, is due to ice. 
 The floors of both Theophilus and Copernicus show ridges that may be lateral 
 moraines. 
 
 What we ordinarily speak of as the lunar day is twenty-nine and a half terrestrial 
 days in length. From the standpoint of climate it may quite as properly be called 
 the lunar year. In the latter case sunrise corresponds to spring, and sunset to 
 autumn. The interval between them is very nearly fifteen of our days. Using the 
 terms in this sense, we find that there are numerous spots scattered over the surface 
 of the Moon which as the season progresses gradually darken. They reach their 
 maximum development about or soon after midsummer, and from thence on slowly 
 fade out and disappear with the approach of autumn. They are widely distributed 
 over the surface, excepting near the poles, but develop most rapidly in the equatorial 
 regions. It is believed that these variable dark spots are due to vegetation. 
 
 One of them is seen in the northern part of Julius Caesar, the large crater just 
 north of the Ariadseus rill, Figure 29. Others are found to the north and east of it. 
 The summer temperature on the Moon is about that of our desert regions at midday. 
 The winter temperature approaches absolute zero, but since grain and other seeds 
 have been exposed without injury to the temperature of liquid air, it seems clear that 
 even terrestrial vegetation can stand a range of temperature quite comparable to that 
 found upon the Moon. 
 
 The central area of Figure 35 represents the crater Eratosthenes taken at the time 
 of full moon. On the photograph it measures an inch and a quarter in diameter, and 
 is on a scale of g-^J-^^, or thirty-two miles to the inch. The central peaks are pure 
 white, and cover an area about one quarter of an inch in diameter. Northeast and 
 northwest of them are two dark spots upon the floor, and southeast of the peaks is a 
 very dark area lying partly on the floor and partly on the inside wall. These spots 
 go through various interesting changes in density as the season progresses, at 
 times entirely disappearing. In one place a slow movement of progression at the 
 rate of four feet per hour has been noted. Outside of the crater, large dark areas 
 are seen, which do not, however, lend themselves so readily to measurement as do 
 those within it. Two dark lines lead away from each of the spots at the base of the 
 central peaks. These lines are believed to be analogous to the so-called canals of 
 Mars. 
 
PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 17? 
 
 Drawings of the Moon and planets show much finer detail than it is possible to 
 
 >tam by any photographs. Figure 37 is a drawing of Eratosthenes made thirty- 
 eight hours later in the lunation than the photograph. Few changes have taken 
 place in the meantime, and the three dark areas within the crater can be readily 
 recognized. It will be noticed that numerous fine canals not at all visible in the 
 photograph appear in the drawing. A much more detailed account of this crater will 
 be found in the Annals of Harvard College Observatory, LIII, 75. 
 
 Since different observers sometimes represent the same detail by different methods 
 of shading, and since to some eyes fine markings, like canals, appear of much less 
 breadth than to others, it is very desirable where two drawings are to be compared, 
 that they should both be, if possible, by the same observer. The sketch of Mars, 
 Figure 38, was made by the writer when at the Lowell Observatory in Arizona. The 
 similarity in appearance of the canals in these two figures, together with their varia- 
 bility under similar conditions, leads one to believe that they are due to the same 
 cause, namely, vegetation. 
 
 During the summer of 1904 the writer was able to spend eight weeks at the 
 Lowe Observatory in southern California. A considerable portion of his time was 
 devoted to a study of Eratosthenes. The interior was found to be seamed by 
 numerous fine cracks. Watching some of these cracks soon after the sun rose upon 
 them, he was able to see them broaden out and change gradually into canals. It 
 is his belief that the cracks gave out water vapor, which fertilized the vegetation 
 along their sides and in their vicinity, and that it was the growth of this vegetation 
 that produced the appearance of a canal. 
 
 The canals of Mars are on a much larger scale than those of the Moon ; one of 
 them indeed reaches the enormous length of 3500 miles. If they are produced 
 naturally, the surface of the planet must be cracked in many places. It is generally 
 thought that terrestrial volcanoes lie along subterranean cracks that do not reach the 
 surface. The volcanoes of the great chain of the Andes lie along a straight crack 
 reaching from southern Peru to Terra del Fuego, 2500 miles in length. The vol- 
 canoes of the Aleutian Islands lie along a curved crack equally long. Since other 
 shorter lines of volcanoes are very numerous upon the Earth, and since countless 
 others existed in former times, the cracks in the Earth's crust must be exceed- 
 ingly numerous. Every dike and mineral vein indeed bears witness to this fact. 
 There is no reason why terrestrial cracks should not be as numerous as those upon 
 the Moon. In the case of the Earth they have usually been closed, sometimes by 
 liquid matter from below, and sometimes by surface denudation. There is one 
 
178 PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES COMPARED. 
 
 crack, however, which comes to the surface in various places in eastern Asia and 
 western Africa, and stretching from the Dead Sea to Lake Nyassa, reaches the 
 enormous length of 3500 miles. The longest known crack upon the Moon, that 
 of Sirsalis, measures about 400 miles. 
 
 It does not necessarily follow, however, even if both the Martian and lunar 
 canals are due to vegetation, that the vegetation is watered in the same manner. 
 There is certainly an atmospheric circulation upon Mars, giving rise to clouds, that 
 might aid materially any subterranean forces. Annals of Harvard College Observa- 
 tory, LIII, 155. Whether these forces could be directed intelligently by assumed 
 intelligent inhabitants of the planet we do not know. The only argument in favor 
 of the existence of such inhabitants is the artificial appearance of the canals. The 
 four canals radiating from the little lake just above the centre of Figure 37 will 
 appear to some minds quite as artificial in appearance as the four canals radiating 
 from the elongated lake to the right of the centre of Figure 38. 
 
 To the southeast and southwest of Kilauea lies a desert region crossed in places 
 by steam cracks. One of the first questions that I asked my guide on reaching the 
 Volcano House was if any of these cracks were still active. He assured me that 
 such was the case, and the next day we visited some that were near Keanakakoi. 
 The desert was found to be absolutely barren except for certain long, narrow strips 
 of vegetation. These consisted chiefly of ferns, some bushes, and a few trees. It 
 was found that they grew over the steam cracks, one of which is shown in Figure 39. 
 This particular crack was a yard in width, over two yards in depth, and about thirty 
 yards long. If the region could have been viewed from a slight elevation we should 
 have found a system of canals crossing the desert, these canals being due to vege- 
 tation, and differing in appearance from those found upon the Moon and Mars in no 
 respect save size. 
 
REFERENCE INDEX TO DESCRIPTIONS OF THE ILLUSTRATIONS. 
 
 SECTIONAL DRAWINGS ON PAGE 171. 
 
 Fig. a. 154, 169, 169. 
 
 " 6. 154, 169, 169. 
 
 c. 155, 169, 169. 
 
 " d. 157, 158, 169, 169. 
 
 e. 155, 169, 169. 
 
 " /. 157, 169. 
 
 " g. 164, 168, 169. 
 
 " h. 164, 168, 169. 
 
 Fig. i. 165, 169. 
 
 " j. 167, 168, 169. 
 
 " A;. 157, 165, 167, 168, 169. 
 
 I 157, 168, 168, 169. 
 
 m. 157, 168, 169. 
 
 " n. 168,169. 
 
 " o. 168, 169, 169. 
 
 " p. 157, 168, 168, 169. 
 
 PHOTOGRAPHS. 
 
 Fig. 1. 154. 
 
 2. 155. 
 
 3. 155, 158, 165, 173. 
 
 4. 156, 158. 
 
 5. 156, 157, 167, 168. 
 
 6. 156, 164, 175. 
 
 7. 157. 
 
 8. 157, 166. 
 
 9. 158, 164. 
 
 10. 158, 158, 164. 
 
 11. 159, 159, 160, 165. 
 
 12. 160, 163, 167. 
 
 13. 160, 163. 
 
 14. 157, 160, 161. 
 
 15. 157, 164. 
 
 16. 161, 164, 167, 168, 170, 170. 
 
 17. 166. 
 
 18. 165. 
 
 19. 166, 166, 168. 
 
 20. 166, 166. 
 
 Fig. 21. 166. 
 
 22. 167. 
 
 23. 170. 
 
 24. 170. 
 
 25. 170. 
 
 " 26. 170. 
 
 " 27. 172. 
 
 28. 172. 
 
 29. 157, 173, 173, 176. 
 
 30. 174. 
 
 31. 157, 172, 173. 
 
 " 32. 161, 174. 
 
 33. 174. 
 
 34. 175. 
 
 35. 157, 176. 
 
 36. 175. 
 
 37. 177, 178. 
 
 38. 177, 178. 
 
 39. 178. 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. I. DIAMOND HEAD 
 
 tf. 
 
 I 
 
 FIG. 2. CINDER CONES ON MAUNA KEA 
 
 H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
OF THE 
 
 I UNIVERSITY ) 
 
 OF 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 3. INTERIOR OF HALEAKALA 
 
 
 
 
 FIG. 4. MAUNA LOA 
 
 \\ 
 
 . H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. 
 
 XIII. 
 
 FIG. 5. BULLIALDUS 
 
 FIG 6. KIES AND MERCATOR 
 
 FIG. 7. KAUHAKU MOLOKAI 
 
 W. H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
OF THE 
 
 UNIVERSITY ) 
 
 OF 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 8. SIXTH CRATER NEAR KILAUEA 
 
 FIG. 9. MOKUAWEOWEO. MAUNA LOA 
 
 W. H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
OF THE 
 
 UNIVERSITY 
 
 OF 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 10. KILAUEA FROM WALDRON S LEDGE 
 
 
 FIG. I I. SLAG SECTION 
 
 W. H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 
 & 
 
 FIG. 12. LAVA LAKE IN KILAUEA 
 
 r 
 
 
 
 FIG. 13 LAVA LAKE IN KILAUEA 
 
 \Y 
 
 H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
OF THE 
 
 VERSITY 
 
 OF 
 
 TOR^ 
 
FIG. 14. SCHICKARD. PHOCYLIDES 
 
 FIG. I 5. SINUS IRIOUM 
 
 
 3fc^ N V ' /' ? ' ' 
 
 FIG. 17. RIDGES. MARE SERENITATIS 
 
 FIGURE 16 is INVERTED. 
 
 vND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 18. MOKUAWEOWEO. MAUNA LOA 
 
 FIG. 19. CRATERLETS KILAUEA IKI 
 
 H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
OF THE 
 
 "NIVERSITY 
 
 OF 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 20. TOP OF CRATERUET. KIUAUEA IKI 
 
 FIG. 21. CRATER PIT. KILAUEA 
 
 w 
 
 . H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
V" OF THE 
 
 UNIVERSITY 
 
 OF 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 22. PART OF CRATER BOWL. HUALALAI 
 
 
 
 FIG. 23. SPIRACLES KILAUEA 
 
 w 
 
 H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG 24. SPIRACLES. KILAUEA 
 
 
 FIG. 25. SPIRACLES. HUALALAI 
 
 vv 
 
 H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
OF THE 
 
 MIVERSITY 
 
 of 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 26. PINNACLE. KILAUEA 
 
 FIG. 27. PINNACLE. HALEAKALA 
 
 W 
 
 H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 ^T\K^; ~>r- 
 -ttvS 
 
 FIG. 28. PINACLES. MARE IMBRIUM 
 
 FIG. 29. ARIADAEUS RILL 
 
 FIG. 30. CRACK. KILAUEA 
 
 W. H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 31. VALLEY OF RHEITA 
 
 FIG. 32. BULLIALDUS 
 
 FIG. 33. CRACK. HUEHUE 
 
 W. H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. XIII. 
 
 FIG. 34. THEOPHILUS 
 
 FIG. 35. ERATOSTHENES 
 
 FIG. 36. EROSION VALLEYS FROM TANTALUS. OAHU 
 
 W. H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
MEMOIRS AMERICAN ACADEMY. VOL. XII. 
 
 Pt 
 
 FIG. 37. LUNAR CANALS 
 
 FIG. 38. MARTIAN CANALS 
 
 FIG. 39. TERRESTRIAL CANAL NEAR KILAUEA 
 
 W. H. PICKERING. LUNAR AND HAWAIIAN PHYSICAL FEATURES 
 
WILL. BE ASSESSED FOR FAILURE TO RETURN 
 THIS BOOK ON THE DATE DUE. THE PENALTY 
 WILL INCREASE TO SO CENTS ON THE FOURTH 
 DAY AND TO $1.OO ON THE SEVENTH DAY 
 OVERDUE. 
 
 JUN 24 ib45 
 
 LD 21-100m-7,'40 (6936s) 
 
179098