Well, this laying is arranged in a most singular fashion. There is first, at a suitable distance from the bottom or the node of the reed, a transversal partition, perpendicular to the axis of the tube. This marks off a cell of unusual size, in which a female is lodged. After that, in view of the excessive width of the tunnel, which is too great for a series in single file, the Osmia appears to alter her mind. She therefore builds a partition perpendicular to the transversal partition which she has just constructed and thus divides the second storey into two rooms, a larger room, in which she lodges a female, and a smaller, in which she lodges a male. She next builds a second transversal partition and a second longitudinal partition perpendicular to it. These once more give two unequal chambers, stocked likewise, the large one with a female, the smaller one with a male. From this third storey onwards, the Osmia abandons geometrical accuracy; the architect seems to be a little out in her reckoning. The transversal partitions become more and more slanting and the work grows irregular, but always with a sprinkling of large chambers for the females and small chambers for the males. Three females and two males are housed in this way, the sexes alternating. By the time that the base of the eleventh cell is reached, the transversal partition is once more almost perpendicular to the axis. Here what happened at the bottom is repeated. There is no longitudinal partition; and the spacious cell, covering the whole diameter of the cylinder, receives a female. The edifice ends with two transversal partitions and one longitudinal partition, which mark out, on the same level, chambers twelve and thirteen, both of which contain males. There is nothing more curious than this mixing of the two sexes, when we know with what precision the Osmia separates them in a linear series, where the narrow width of the cylinder demands that the cells shall be set singly, one above the other. Here, the Bee is making use of a tube whose diameter is not suited to her work; she is constructing a complex and difficult edifice, which perhaps would not possess the necessary solidity if the ceilings were too broad. The Osmia therefore supports these ceilings with longitudinal partitions; and the unequal chambers resulting from the introduction of these partitions receive females at one time and males at another, according to their capacity. CHAPTER 5. PERMUTATIONS OF SEX.