is that often taken by Octopus Bairdii when resting on the bottom after swimming, with the head raised, and the body supported on the thickest part of the arms. The ends of the arms are curved irregularly, as they might be in an animal just starting to crawl.

The highest point of the body is twenty-two inches above the lowest suckers. The arms spread over a circle eighteen feet in diameter, and the connecting membrane between the lateral arms extends three feet from the mouth. The longest arms, those of the second pair, are made as long as the largest measurements from life (sixteen feet); and the shortest, the fourth pair, thirteen feet. The third arm on the right side is shorter than the others, and hectocotylized in the male, and is so made in the model. All the arms are four inches in diameter at the thickest part. The body is made proportionally smaller than in small specimens. The warts on the head are copied from one of the largest specimens examined, the others showing only two pairs over the eyes. The membranes between the arms have been made much as they are in alcohol, but somewhat wider and more distinct along the sides of the arms. The largest suckers are two and a quarter inches in diameter, and decrease in size from the thickest part of the arm toward the tip, and toward the mouth.

For convenience in making and moving the model, the arms are made removable at a distance of three feet from the mouth, just beyond the edge of the widest membrane.

The upper side of the middle part of the model, including the head and body, was modelled in clay, and a mould made from it in plaster. This was then turned over, and the mouth and under sides of the bases of the arms modelled in it. The arms are so much alike that it was only necessary to model the bases of two of them, one right and one left; and from these a plaster mould was taken in which the casts of the bases of all the arms were made. This mould stands against the table at the left in the engraving. The ends of the arms were modelled in a similar way, the back being first finished, and a plaster mould made, which was turned over, and the under side modelled upon it. For modelling For modelling the tops of the suckers, a set of stamps was made. A set of suckers of the desired sizes was modelled in clay on a turntable, and plaster casts made of the tops of them, and these used to stamp the tops of the rest of the suckers, which were trimmed round with a knife, and attached to the arm with soft clay, after which, the narrow membranes connect

ing the larger suckers were modelled between them.

When the moulds were dry, the paper casts were made in them by methods which had been used by Mr. Palmer for models of large fishes and cetacea. The moulds having been greased, paper soaked with paste was laid in it, and pressed and rubbed with the hands until it fitted close to the surface of the mould, and the edges of the pieces of paper adhered together. When the first layer of paper was nearly dry, another was pasted over it; and, if the strength of the model required it, other layers were added. The thin membranes between the arms were strengthened by wire netting between the two layers of paper, the meshes being filled with whiting mixed with glue. On the surfaces of the suckers, paper pulp was put in the mould before the paper was pasted in.

After drying several days, the casts were taken from the moulds, the edges trimmed, and the pieces fastened together with glue. The broken places in the casts were mended with paper pulp, the joints covered over with the same material, and, when dry, the surface was smoothed with sandpaper, and varnished with shellac. The siphon was made separately, and afterward attached to the body. The mouth was made of plaster, showing the jaws closed. The eyes are of glass, like ordinary birds'-eyes, painted and silvered according to the best evidence we could get as to their color.

The color of Octopus punctatus seems to differ greatly, according to its moods and surroundings. It is commonly described as light orange or yellow with reddish-brown spots. At other times it appears to be bright orange and crimson, with dark-brown blotches on the back. The model was first painted light gray, on which the other colors were thrown from a brush in fine spots. The orange spots are scattered over the whole surface, and more thickly in patches along the back and sides of the arms. Crimson spots are distributed in the same way; and over both, dark-brown spots are thinly scattered. The faces of the suckers are yellowish white without spots.

The model weighs about seventy pounds, and is stiff and strong enough for ordinary handling, and only liable to be broken by a fall or sudden blow. It is intended to be hung in a horizontal position, as in the engraving, but high enough for the under side to be seen, as well as the upper. It hangs by eight wires attached to rings near the joints in the arms, and connected together above so that it can be hung from a single hook.

The engraving shows the model hanging in

2.50

2.40

2.30

2.20

liam H. Brewer, on The evolution of the American trottinghorse. He remarks, "The formation of this new breed," for such he considers it, "is so recent, the development of a special quality has been so marked, there is such an abundant literature pertaining to its history, the system of sporting records is SO

1818

3.00

1852

2.26

1871

2.17

1824

2.40

1853

2.25

1872

2.16

66

2.34

1856

Looking at the rapidly diminishing numbers of the record' columns of table I., one cannot but ask, When will this diminution cease? When will the fastest time attainable be reached? Plotting records as ordinates, and dates as abscissas, one would naturally expect the first dy differential, da' to diminish in value, and that the curve would be asymptotic to the axis of x : or, in other words, we should expect, as time went on, that the speed of the horse would not improve so rapidly as it did at first, and that after a while the improvement would be so slow that we might consider that we had prac

time at all. So that our curve must sooner or later become a straight line, and ultimately concave upwards. Drawing a straight line which shall agree as nearly as possible with our observations, we shall find from it, that the speed of the trotting-horse is increasing at a nearly uniform rate of 43 seconds in ten years; so that, on this supposition, it would cross the two-minute line in 1907, and the one-minute line in 2045. It is highly probable that the curve will have become concave before the latter period; but it does not seem too rash to predict that a horse will be born before 1907 that can trot a mile in two minutes.

II. — Total number of horses capable of trotting

slope, which must have existed at least up to the end of the carboniferous, was then or soon after reversed in the slow writhing of the surface. This is demanded by the lay of the land, and by the now small area of what must have been, in paleozoic time, a large crystalline landmass. The slope being changed early in the growth of the folds, or before their beginning, the streams tried to make their way to the eastward; and the Hudson, Delaware, Susquehanna, Potomac, and James are the descendants of those that succeeded. Their rectangular courses, alternately longitudinal and transverse, bear witness to their defeats and victories. Lakes must have been numerous here once, though they are now all drained. It is known that rivers often chose cross-faults of small throw as points of attack in cutting their way through the growing ridges; and it is very probable that they made use of pre-existent valleys when they advanced over the old sinking land.

In considering the applicability of backwardcutting lateral streams to the production of our cross-valleys, we should test the past by the present, and examine such ridges as Kittatinny or Bald Eagle mountains in Pennsylvania, or Clinch mountain in Tennessee, rising between parallel longitudinal valleys, to see if they show embryonic cross-valleys in the more advanced stages of development. They do not. The continuity of their crest-line is most characteristic and remarkable: it very rarely departs from its line of almost uniform height. The exceptions are, first, the finished watergaps, or transverse valleys, whose origin is in discussion; second, the occasional wind-gaps, or notches, which sometimes cut the ridge a third or half way to its base, and which are, we believe, always determined by small transverse faults; third, the less conspicuous serrations of small value. It is difficult to assign any reason why lateral streams should not now, as well as in former times, show us the later 'stages of breaking down the ridge on which they rise; and yet these almost-formed crossvalleys between adjoining longitudinal valleys are practically unknown in our Appalachian topography. The reason of their absence can hardly be, that there are now enough completed water-gaps for all practical purposes, and hence the lateral streams stop making any more; for this would imply a consciousness of the end that plays no part in geological operations, and we are therefore constrained to think that Löwl's explanation cannot apply to the Appalachians in any general way.

But it has a certain limited application in

the making of coves,' as may be perceived by the following considerations. The backwardcutting of a lateral stream can form a water-gap only where the longitudinal valley into which the lateral stream flows is decidedly lower than the longitudinal valley on the other side of the dividing-ridge; for, if there is no such difference of level, the pass through the ridge between the two will be eroded more and more slowly as it is lowered, and finally it will remain at practically a constant altitude above the valleys on either side. It can never form a drainage channel joining them: but, if the longitudinal valleys are of different heights, the result as described by Löwl may be produced; or, if a broad plateau-fold is bordered by a deep valley, its lateral streams may finally head up in coves or circular valleys, like those south of the west branch of the Susquehanna, and at many other points in the Appalachians. The geological map of Pennsylvania (1858) shows these admirably.

6

Now it may be asked, Are not the upper valleys of the Susquehanna, and of the other rivers that break through the Blue mountain, merely large examples of coves'? There are two objections to this explanation. First, what became of the head waters of these rivers before they had a south-easterly outlet? It seems most probable, that the many pre-existent streams in each river-basin concentrated their waters in a single channel of overflow, and that this one channel survives, a fine example of natural selection. Second, how does it happen- notably in the case of the Susquehanna just above Harrisburg- that several deep water-gaps have been formed one behind the other? Such an arrangement might naturally result if the valley were antecedent; but it is difficult to account for if the several gaps result from the backward erosion of accidental lateral streams.

Löwl thinks that faults are greater obstacles to rivers than folds. He says, that even if river erosion could, under certain favorable conditions, keep pace with mountain folding, it does not follow that it could control a fault: for that would imply that the fault was formed gradually, and that its throw increased at a constant though imperceptible rate; and this he considers entirely unwarranted (408). It is certainly a difficult matter to understand the mechanics of such faults; and yet our ideas concerning them must conform to the facts as they occur in nature. In spite, therefore, of a natural preference for an active growth of faults, we are compelled, when we see streams running across them from the downthrow to the

upthrow side, to accord them a slow growth. Tennessee shows many examples of this paradoxical nature; and some of the faults thus disregarded, or, we might say, corrected, have a throw of several thousand feet. On the other hand, it cannot be denied that many faults have had a controlling influence on stream-courses ; and we must therefore admit here, as above, the possibility of valley-cutting being stronger or weaker than orographic movements. The variety that is to be seen in the physical features of the earth, and that is consequently to be looked for in the conditions which determined them, is so great that it demands almost equal variety in the theories for their explanation. W. M. DAVIS.

Cambridge, Jan. 12, 1883.

THE ORIGIN, AFTER BIRTH, OF ASPIRATION OF THE THORAX.

THE negative pressure in the pleural cavities, which plays such an important part in the respiratory mechanism of the adult mammal, and also exerts a marked influence upon the flow of blood and lymph, is known not to exist in unborn or stillborn mammals which have never breathed. It has hitherto been assumed, however, that it was established with the first inspiration; and all theories as to its mode of production have been controlled by this belief. Hermann, in an interesting paper (Pflüg. archiv, xxx. 276), shows that this assumption is incorrect. In infants which have lived and breathed after birth for periods of from one hour to eight days, there is found on experiment to be no negative pressure in the pleural cavity when the chest-walls are in their death position. This fact leads, necessarily, to important results in regard to the respiration of young mammals. Their lungs in expiration can contain hardly any of what in the adult is known as 'residual' air: they contain still some air imprisoned in the air-cells, and causing them to float in water. But this minimal'

air is practically no more than what remains in a piece of adult lung squeezed between thumb and finger. Except this minute quantity, there is in the new-born mammal no stationary air. At each inspiration, air direct from outside enters the alveoli of the lungs; and, at each expiration, air from the alveoli is expelled, leaving the lungs practically empty. Hence the renewal of the air in the lungs is much more efficient than in adults. The high percentage of oxygen which the alveolar air must contain is probably correlated with the more active oxidations known to occur in the young animal. The question naturally arises, What is the object of the residual air in the lungs, and the negative pressure in the thorax, which we find established later? Hermann suggests three possible objects: 1°. Aspiration on the veins, promoting blood flow to the heart; 2°. More uniform composition of air in the alveoli [and, we may add, more uniform temperature]; 3°. The presence of a certain store of air in the lungs in the case of a temporary stoppage of the breathing-movements. It remains to be seen at what age and rate the negative pressure in the thorax is developed. It is obviously brought about by a more rapid growth of the thorax than of the lungs. H. NEWELL MARTIN.

PLANT-LIFE, PAST AND PRESENT.

THE opening lecture of the second course of 'Saturday lectures,' delivered at the National museum in Washington, was by Mr. Lester F. Ward, assistant geologist U. S. geological survey, and honorary curator of fossil plants to the museum; the subject being 'Plant-life of the globe, past and present.'

The object of the lecture was to give some account of the progress which has taken place toward the adoption of a truly natural system of botanical classification. After describing and comparing the methods of Linné, of A. L. de Jussieu, of Adrien de Jussieu, and of modern botanists, the lecturer pointed out the objections which may be made to all of these, and then presented the outline of a system which aimed to exclude the objectionable features, and to accord with the results of the latest discoveries in structural botany, and especially with the teachings of paleontology, which he claimed to have been too much ignored by botanists. The proposed system was as follows:

type which have been found fossil at each geological horizon, and also the most reliable estimates that could be obtained of the number living at the present time in all parts of the world. It also showed the percentage that each type formed of the total known flora of each epoch. We give below a condensed view of this chart, which is all we have space to present.

Relative to this table, it should be explained,

1. That the figures given for the living gymnosperms and dicotyledons are, in round numbers, those of Messrs. Bentham and Hooker, as stated for each genus and order in the Genera plantarum,' and which are here compiled, perhaps for the first time.

2. That the number of fossil species were collated from a great number of sources; Schimper's Traité de paléontologie végétale' being the basis, supplemented by data from all the more recent publications which were accessible, and by some unpublished data. Absolute completeness, however, was not claimed, but only such substantial accuracy as was deemed sufficient for the purposes of the lecture.

3. That under tertiary time' are included all the beds from the quaternary to the middle cretaceous; the latter being represented in this country by the Dakota group, and in Europe by the cenomanian. This is done because it is at the last-named horizon that the dicotyledons first appear, and because they appear here in such extraordinary profusion. Marquis Saporta has also made the vegetable tertiary to begin at this point.1

The facts embodied in this table were further graphically illustrated by two diagrams, prepared by Ensign E. E. Hayden, U.S.N. The first of these showed, by means of accurately plotted curves and

CRYPTOGAMS.

VASCULAR.

CELLULAR.

Cycadaceae.

Coniferac

Gnetaccae.

Monocotyle

dons.

75

0.04

300 0.20

The claims of this scheme as the nearest approach yet made to the system of nature were supported, for the most part, on paleontological grounds. To do this, an elaborate chart was presented, giving the geological history of each of the principal types of vegetation. This was in the form of a tabular exhibit of the number of species belonging to each

colored areas, the development of each type of vegetation through the several ascending strata, the breadth of the areas at any epoch representing the prominence of the several types relatively to the entire flora of that epoch. The other diagram consisted of

1 Le monde des plantes avant l'apparition de l'homme, p. 160.

25

94

1,186