beneath the projecting edge and pry it up by bending the first joint of the finger; the resistance is scarcely perceptible.
On the other side of the veranda, once the eye is accustomed to the darkness, a paler form can be seen outlined against the wall of the house: Franck's white shirt. His forearms are lying on the elbow-rests. The upper part of his body is leaning back in the chair.
A . . . is humming a dance tune whose words remain unintelligible. But perhaps Franck understands them, if he already knows them, from having heard them often, perhaps with her. Perhaps it is one of her favorite records.
A . . .’s arms, a little less distinct than her neighbor's because of the color—though light—of the material of her dress, are also lying on the elbow-rests of her chair. The four hands are lying in a row, motionless. The space between A . . .’s left hand and Franck's right hand is approximately two inches. The shrill cry of some nocturnal carnivore, sharp and short, echoes again toward the bottom of the valley, at an unspecifiable distance.
“I think I'll be getting along,” Franck says.
“Oh don't go,” A . . . replies at once, “it's not late at all. It's so pleasant sitting out here.”
If Franck wanted to leave, he would have a good excuse: his wife and child who are alone in the house. But he mentions only the hour he must get up the next morning, without making any reference to Christiane. The same shrill, short cry, which sounds closer, now seems to come from the garden, quite near the foot of the veranda on the east side.
As if echoing it, a similar cry follows, coming from the opposite direction. Others answer these, from higher up, toward the road; then still others, from the low ground.
Sometimes the sound is a little lower, or more prolonged. There are probably different kinds of animals. Still, all these cries are alike; not that their common characteristic is easy to decide, but rather their common lack of characteristics: they do not seem to be cries of fright, or pain, or intimidation, or even love. They sound like mechanical cries, uttered without perceptible motive, expressing nothing, indicating only the existence, the position, and the respective movements of each animal, whose trajectory through the night they punctuate.
“All the same,” Franck says, “I think I'll be getting along.”
A . . . does not reply. Neither one has moved. They are sitting side by side, leaning back in their chairs, arms lying on the elbow-rests, their four hands in similar positions, at the same level, lined up parallel to the wall of the house.
Now the shadow of the southwest column—at the corner of the veranda on the bedroom side—falls across the garden. The sun, still low in the eastern sky, rakes the valley from the side. The rows of banana trees, growing at an angle to the direction of the valley, are everywhere quite distinct in this light.
From the bottom to the upper edge of the highest sectors, on the hillside facing the one the house is built on, it is relatively easy to count the trees; particularly opposite the house, thanks to the recent plantings of the patches located in this area.
The valley has been cleared over the greater part of its width here: there remains, at present, nothing but a border of brush (some thirty yards across at the top of the plateau) which joins the valley by a knoll with neither crest nor rocky fall.
The line of separation between the uncultivated zone and the banana plantation is not entirely straight. It is a zigzag line, with alternately protruding and receding angles, each belonging to a different patch of different age, but of a generally identical orientation.
Just opposite the house, a clump of trees marks the highest point the cultivation reaches in this sector. The patch that ends here is a rectangle. The ground is invisible, or virtually so, between the fronds. Still, the impeccable alignment of the boles shows that they have been planted only recently and that no stems have as yet been cut.
Starting from this clump of trees, the patch runs downhill with a slight divergence (toward the left) from the greatest angle of slope. There are thirty-two banana trees in the row, down to the lower edge of the patch.
Prolonging this patch toward the bottom, with the same arrangement of rows, another patch occupies the space included between the first patch and the little stream that flows through the valley bottom. This second patch is twenty-three trees deep, and only its more advanced vegetation distinguishes it from the preceding patch: the greater height of the trunks, the tangle of fronds, and the number of well-formed steins. Besides, some stems have already been cut. But the empty place where the bole has been cut is then as easily discernible as the tree itself would be with its tuft of wide, pale-green leaves, out of which comes the thick curving stem bearing the fruit.
Furthermore, instead of being rectangular like the one above it, this patch is trapezoidal; for the stream bank that constitutes its lower edge is not perpendicular to its two sides—running up the slope—which are parallel to each other. The row on the right side has no more than thirteen banana trees instead of twenty-three.
And finally, the lower edge of this patch is not straight, since the little stream is not: a slight bulge narrows the patch toward the middle of its width. The central row, which should have eighteen trees if it were to be a true trapezoid, has, in fact, only sixteen.
In the second row, starting from the far left, there would be twenty-two trees (because of the alternate arrangement) in the case of a rectangular patch. There would also be twenty-two lor a patch that was precisely trapezoidal, the reduction being scarcely noticeable at such a short distance from its base. And, in fact, there are twenty-two trees there.
But the third row too has only twenty-two trees, instead of twenty-three which the alternately-arranged rectangle would have. No additional difference is introduced, at this level, by the bulge in the lower edge. The same is true for the fourth row, which includes twenty-one boles, that is, one less than an even row of the imaginary rectangle.
The bulge of the bank also begins to take effect starting from the fifth row: this row, as a matter of fact, also possesses only twenty-one trees, whereas it should have twenty-two for a true trapezoid and twenty-three for a rectangle (uneven row).
These numbers themselves are theoretical, since certain banana trees have already been cut at ground level, once the stem has matured. There are actually nineteen tufts of leaves and two empty spaces which constitute the fourth row; and in the fifth, twenty tufts and one space—that is, from bottom to top: eight tufts of leaves, an empty space, twelve tufts of leaves.
Without bothering with the order in which the actually visible banana trees and the cut banana trees occur, the sixth row gives the following numbers: twenty-two, twenty-one, twenty, nineteen—which represent respectively the rectangle, the true trapezoid, the trapezoid with a curved edge, and the same after subtracting the boles cut for the harvest.
And for the following rows: twenty-three, twenty-one, twenty-one, twenty-one. Twenty-two, twenty-one, twenty, twenty. Twenty-three, twenty-one, twenty, nineteen, etc . . . .
On the log bridge that crosses the stream at the bottom edge of this patch, there is a man crouching: a native, wearing blue trousers and a colorless undershirt that leaves his shoulders bare. He is leaning toward the liquid surface, as if he were trying to see something at the bottom, which is scarcely possible, the water never being transparent enough despite its extreme shallowness.
On the near slope of the valley, a single patch runs up-hill from the stream to the garden. Despite the rather slight declivity the slope appears to have, the banana trees are still easy to count here from the height of the veranda. As a matter of fact, the trees are very young in this zone, which has only recently been replanted. Not only is the regularity of the planting perfect here, but the trunks are no more than a foot and a half high, and the tufts of leaves that terminate them are still quite far apart from each other. Finally, the angle of the rows with the direction of the valley (about forty-five degrees) also favors their enumeration.
An oblique row begins at the log bridge, at the right, and reaches the left corner of the garden. It includes the thirty-six trees in its length. The alternate arrangement makes it possible to consider these