I have been studying the bramble-dwellers for seven or eight years and I could not say how many strings of coc.o.o.ns have pa.s.sed through my hands.

During a recent winter, in view particularly of the distribution of the s.e.xes, I collected some forty of this Osmia's nests, transferred their contents into gla.s.s tubes and made a careful summary of the s.e.xes.

I give some of my results. The figures start in their order from the bottom of the tunnel dug in the bramble and proceed upwards to the orifice. The figure 1 therefore denotes the first-born of the series, the oldest in date; the highest figure denotes the last-born. The letter M, placed under the corresponding figure, represents the male and the letter F the female s.e.x.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 F F M F M F M M F F F F M F M

This is the longest series that I have ever been able to procure. It is also complete, inasmuch as it comprises the entire laying of the Osmia.

My statement requires explaining, otherwise it would seem impossible to know whether a mother whose acts one has not watched, nay more, whom one has never seen, has or has not finished laying her eggs. The bramble-stump under consideration leaves a free s.p.a.ce of nearly four inches above the continuous string of coc.o.o.ns. Beyond it, at the actual orifice, is the terminal stopper, the thick plug which closes the entrance to the gallery. In this empty portion of the tunnel there is ample accommodation for numerous coc.o.o.ns. The fact that the mother has not made use of it proves that her ovaries were exhausted; for it is exceedingly unlikely that she has abandoned first-rate lodgings to go laboriously digging a new gallery elsewhere and there continue her laying.

You may say that, if the unoccupied s.p.a.ce marks the end of the laying, nothing tells us that the beginning is actually at the bottom of the cul-de-sac, at the other end of the tunnel. You may also say that the laying is done in shifts, separated by intervals of rest. The s.p.a.ce left empty in the channel would mean that one of these shifts was finished and not that there were no more eggs ripe for hatching. In answer to these very plausible explanations, I will say that, the sum of my observations--and they have been extremely numerous--is that the total number of eggs laid not only by the Osmiae but by a host of other Bees fluctuates round about fifteen.

Besides, when we consider that the active life of these insects lasts hardly a month; when we remember that this period of activity is disturbed by dark, rainy or very windy days, during which all work is suspended; when lastly we ascertain, as I have done ad nauseam in the case of the Three-horned Osmia, the time required for building and victualling a cell, it becomes obvious that the total laying must be kept within narrow bounds and that the mother has no time to lose if she wishes to get fifteen cells satisfactorily built in three or four weeks interrupted by compulsory rests. I shall give some facts later which will dispel your doubts, if any remain.

I a.s.sume, therefore, that a number of eggs bordering on fifteen represents the entire family of an Osmia, as it does of many other Bees.

Let us consult some other complete series. Here are two:

1 2 3 4 5 6 7 8 9 10 11 12 13 F F M F M F M F F F F M F F M F F F M F F M F M

In both cases, the laying is taken as complete, for the same reasons as above.

We will end with some series that appear to me incomplete, in view of the small number of cells and the absence of any free s.p.a.ce above the pile of coc.o.o.ns:

1 2 3 4 5 6 7 8 M M F M M M M M M M F M F M M M F M F F M M M M M F M F F F F M M M F M

These examples are more than sufficient. It is quite evident that the distribution of the s.e.xes is not governed by any rule. All that I can say on consulting the whole of my notes, which contain a good many instances of complete layings--most of them, unfortunately, spoilt through gaps caused by parasites, the death of the larva, the failure of the egg to hatch and other accidents--all that I can say in general is that the complete series begins with females and nearly always ends with males. The incomplete series can teach us nothing in this respect, for they are only fragments starting we know not whence; and it is impossible to tell whether they should be ascribed to the beginning, to the end, or to an intermediate period of the laying. To sum up: in the laying of the Three-p.r.o.nged Osmia, no order governs the succession of the s.e.xes; only, the series has a marked tendency to begin with females and to finish with males.

The brambles, in my district, harbour two other Osmiae, both of much smaller size: O. detrita, PEREZ, and O. parvula, DUF. The first is very common, the second very rare; and until now I have found only one of her nests, placed above a nest of O. detrita, in the same bramble. Here, instead of the lack of order in the distribution of the s.e.xes which we find with O. tridentata, we have an order remarkable for consistency and simplicity. I have before me the list of the series of O. detrita collected last winter. Here are some of them:

1. A series of twelve: seven females, beginning with the bottom of the tunnel, and then five males.

2. A series of nine: three females first, then six males.

3. A series of eight: five females followed by three males.

4. A series of eight: seven females followed by one male.

5. A series of eight: one female followed by seven males.

6. A series of seven: six females followed by one male.

The first series might very well be complete. The second and fifth appear to be the end of layings, of which the beginning has taken place elsewhere, in another bramble-stump. The males predominate and finish off the series. Nos. 3, 4 and 6, on the other hand, look like the beginnings of layings: the females predominate and are at the head of the series. Even if these interpretations should be open to doubt, one result at least is certain: with O. detrita, the laying is divided into two groups, with no intermingling of the s.e.xes; the first group laid yields nothing but females, the second, or more recent, yields nothing but males.

What was only a sort of attempt with the Three-p.r.o.nged Osmia--who, it is true, begins with females and ends with males, but muddles up the order and mixes the two s.e.xes anyhow between the extreme points--becomes a regular law with her kinswoman. The mother occupies herself at the start with the stronger s.e.x, the more necessary, the better-gifted, the female s.e.x, to which she devotes the first flush of her laying and the fullness of her vigour; later, when she is perhaps already at the end of her strength, she bestows what remains of her maternal solicitude upon the weaker s.e.x, the less-gifted, almost negligible male s.e.x.

O. parvula, of whom I unfortunately possess but one series, repeats what the previous witness has just shown us. This series, one of nine coc.o.o.ns, comprises five females followed by four males, without any mixing of the s.e.xes.

Next to these disgorgers of honey and gleaners of pollen-dust, it would be well to consult other Hymenoptera, Wasps who devote themselves to the chase and pile their cells one after the other, in a row, showing the relative age of the coc.o.o.ns. The brambles house several of these: Solenius vagus, who stores up Flies; Psen atratus, who provides her grubs with a heap of Plant-lice; Trypoxylon figulus, who feeds them with Spiders.

Solenius vagus digs her gallery in a bramble-stick that is lopped short, but still fresh and green. The house of this Fly-huntress, therefore, suffers from damp, as the sap enters, especially on the lower floors.

This seems to me rather insanitary. To avoid the humidity, or for other reasons which escape me, the Solenius does not dig very far into her bramble-stump and consequently can stack but a small number of cells in it. A series of five coc.o.o.ns gives me first four females and then one male; another series, also of five, contains first three females, with two males following. These are the most complete that I have for the moment.

I reckoned on the Black Psen, or Psen atratus, whose series are pretty long; it is a pity that they are nearly always greatly interfered with by a parasite called Ephialtes mediator. (Cf. "The Life of the Fly": chapter 2.--Translator's Note.) I obtained only three series free from gaps: one of eight coc.o.o.ns, comprising only females; one of six, likewise consisting wholly of females; lastly, one of eight, formed exclusively of males. These instances seem to show that the Psen arranges her laying in a succession of females and a succession of males; but they tell us nothing of the relative order of the two series.

From the Spider-huntress, Trypoxylon figulus, I learnt nothing decisive. She appeared to me to rove about from one bramble to the next, utilizing galleries which she has not dug herself. Not troubling to be economical with a lodging which it has cost her nothing to acquire, she carelessly builds a few part.i.tions at very unequal heights, stuffs three or four compartments with Spiders and pa.s.ses on to another bramble-stump, with no reason, so far as I know, for abandoning the first. Her cells, therefore, occur in series that are too short to give us any useful information.

This is all that the bramble-dwellers have to tell us; I have enumerated the list of the princ.i.p.al ones in my district. We will now look into some other Bees who arrange their coc.o.o.ns in single files: the Megachiles (Cf. Chapter 8 of the present volume.--Translator's Note.), who cut disks out of leaves and fashion the disks into thimble-shaped receptacles; the Anthidia (Cf. Chapters 9 and 10 of the present volume.--Translator's Note.), who weave their honey-wallets out of cotton-wool and arrange their cells one after the other in some cylindrical gallery. In most cases, the home is the produce of neither the one nor the other. A tunnel in the upright, earthy banks, the old work of some Anthophora, is the usual dwelling. There is no great depth to these retreats; and all my searches, zealously prosecuted during a number of winters, procured me only series containing a small number of coc.o.o.ns, four or five at most, often one alone. And, what is quite as serious, nearly all these series are spoilt by parasites and allow me to draw no well-founded deductions.

I remembered finding, at rare intervals, nests of both the Anthidium and the Megachile in the hollows of cut reeds. I thereupon installed some hives of a new kind on the sunniest walls of my enclosure. They consisted of stumps of the great reed of the south, open at one end, closed at the other by the natural knot and gathered into a sort of enormous pan-pipe, such as Polyphemus might have employed. The invitation was accepted: Osmiae, Anthidia and Megachiles came in fairly large numbers, especially the first, to benefit by the queer installation.

In this way I obtained some magnificent series of Anthidia and Megachiles, running up to a dozen. There was a melancholy side to this success. All my series, with not one exception, were ravaged by parasites. Those of the Megachile (M. sericans, FONSCOL), who fashions her goblets with robinia-, holm-, and terebinth-leaves, were inhabited by Coelioxys octodentata (A Parasitic Bee.--Translator's Note.); those of the Anthidium (A. florentinum, LATR.) were occupied by a Leucopsis.

Both kinds were swarming with a colony of pigmy parasites whose name I have not yet been able to discover. In short, my pan-pipe hives, though very useful to me from other points of view, taught me nothing about the order of the s.e.xes among the Leaf-cutters and the cotton-weavers.

I was more fortunate with three Osmiae (O. tricornis, LATR., O. cornuta, LATR., and O. Latreillii, SPIN.), all of whom gave me splendid results, with reed-stumps arranged either against the walls of my garden, as I have just said, or near their customary abode, the huge nests of the Mason-bee of the Sheds. One of them, the Three-horned Osmia, did better still: as I have described, she built her nests in my study, as plentifully as I could wish, using reeds, gla.s.s tubes and other retreats of my selecting for her galleries.

We will consult this last, who has furnished me with doc.u.ments beyond my fondest hopes, and begin by asking her of how many eggs her average laying consists. Of the whole heap of colonized tubes in my study, or else out of doors, in the hurdle-reeds and the pan-pipe appliances, the best-filled contains fifteen cells, with a free s.p.a.ce above the series, a s.p.a.ce showing that the laying is ended, for, if the mother had any more eggs available, she would have lodged them in the room which she leaves unoccupied. This string of fifteen appears to be rare; it was the only one that I found. My attempts at indoor rearing, pursued during two years with gla.s.s tubes or reeds, taught me that the Three-horned Osmia is not much addicted to long series. As though to decrease the difficulties of the coming deliverance, she prefers short galleries, in which only a part of the laying is stacked. We must then follow the same mother in her migration from one dwelling to the next if we would obtain a complete census of her family. A spot of colour, dropped on the Bee's thorax with a paint-brush while she is absorbed in closing up the mouth of the tunnel, enables us to recognize the Osmia in her various homes.

In this way, the swarm that resided in my study furnished me, in the first year, with an average of twelve cells. Next year, the summer appeared to be more favourable and the average became rather higher, reaching fifteen. The most numerous laying performed under my eyes, not in a tube, but in a succession of Snail-sh.e.l.ls, reached the figure of twenty-six. On the other hand, layings of between eight and ten are not uncommon. Lastly, taking all my records together, the result is that the family of the Osmia fluctuates round about fifteen in number.

I have already spoken of the great differences in size apparent in the cells of one and the same series. The part.i.tions, at first widely s.p.a.ced, draw gradually nearer to one another as they come closer to the aperture, which implies roomy cells at the back and narrow cells in front. The contents of these compartments are no less uneven between one portion and another of the string. Without any exception known to me, the large cells, those with which the series starts, have more abundant provisions than the straitened cells with which the series ends. The heap of honey and pollen in the first is twice or even thrice as large as that in the second. In the last cells, the most recent in date, the victuals are but a pinch of pollen, so n.i.g.g.ardly in amount that we wonder what will become of the larva with that meagre ration.

One would think that the Osmia, when nearing the end of the laying, attaches no importance to her last-born, to whom she doles out s.p.a.ce and food so sparingly. The first-born receive the benefit of her early enthusiasm: theirs is the well-spread table, theirs the s.p.a.cious apartments. The work has begun to pall by the time that the last eggs are laid; and the last-comers have to put up with a scurvy portion of food and a tiny corner.

The difference shows itself in another way after the coc.o.o.ns are spun.

The large cells, those at the back, receive the bulky coc.o.o.ns; the small ones, those in front, have coc.o.o.ns only a half or a third as big. Before opening them and ascertaining the s.e.x of the Osmia inside, let us wait for the transformation into the perfect insect, which will take place towards the end of summer. If impatience gets the better of us, we can open them at the end of July or in August. The insect is then in the nymphal stage; and it is easy, under this form, to distinguish the two s.e.xes by the length of the antennae, which are larger in the males, and by the gla.s.sy protuberances on the forehead, the sign of the future armour of the females. Well, the small coc.o.o.ns, those in the narrow front cells, with their scanty store of provisions, all belong to males; the big coc.o.o.ns, those in the s.p.a.cious and well-stocked cells at the back, all belong to females.

The conclusion is definite: the laying of the Three-horned Osmia consists of two distinct groups, first a group of females and then a group of males.

With my pan-pipe apparatus displayed on the walls of my enclosure and with old hurdle-reeds left lying flat out of doors, I obtained the Horned Osmia in fair quant.i.ties. I persuaded Latreille's Osmia to build her nest in reeds, which she did with a zeal which I was far from expecting. All that I had to do was to lay some reed-stumps horizontally within her reach, in the immediate neighbourhood of her usual haunts, namely, the nests of the Mason-bee of the Sheds. Lastly, I succeeded without difficulty in making her build her nests in the privacy of my study, with gla.s.s tubes for a house. The result surpa.s.sed my hopes.

With both these Osmiae, the division of the gallery is the same as with the Three-horned Osmia. At the back are large cells with plentiful provisions and widely-s.p.a.ced part.i.tions; in front, small cells, with scanty provisions and part.i.tions close together. Also, the larger cells supplied me with big coc.o.o.ns and females; the smaller cells gave me little coc.o.o.ns and males. The conclusion therefore is exactly the same in the case of all three Osmiae.

Before dismissing the Osmiae, let us devote a moment to their coc.o.o.ns, a comparison of which, in the matter of bulk, will furnish us with fairly accurate evidence as to the relative size of the two s.e.xes, for the thing contained, the perfect insect, is evidently proportionate to the silken wrapper in which it is enclosed. These coc.o.o.ns are oval-shaped and may be regarded as ellipsoids formed by a revolution around the major axis. The volume of one of these solids is expressed in the following formula:

4 / 3 x pi x a x (b squared),

in which 2a is the major axis and 2b the minor axis.

Now, the average dimensions of the coc.o.o.ns of the Three-horned Osmia are as follows:

2a = 13 mm. (.507 inch.--Translator's Note.), 2b = 7 mm. (.273 inch.--Translator's Note.) in the females;

2a = 9 mm. (.351 inch.--Translator's Note.), 2b = 5 mm. (.195 inch.--Translator's Note.) in the males.

The ratio therefore between 13 x 7 x 7 = 637 and 9 x 5 x 5 = 225 will be more or less the ratio between the sizes of the two s.e.xes. This ratio is somewhere between 2 to 1 and 3 to 1. The females therefore are two or three times larger than the males, a proportion already suggested by a comparison of the ma.s.s of provisions, estimated simply by the eye.

The Horned Osmia gives us the following average dimensions:

2a = 15 mm. (.585 inch.--Translator's Note.), 2b = 9 mm. (.351 inch.--Translator's Note.) in the females;