which has its leaves distichous on the lateral boughs, so as to expose their upper surfaces to the light from above, will produce vertical shoots from the top of the bulb, having their leaves arranged on the plan. The same occurs with the free branches of the ivy compared with those climbing up a wall, on which the leaves are distichous.

The Yew seems to have its leaves distichous, but a closer inspection reveals the fact they are not so, but on the plan: each leaf, however, is twisted at its base, so as to make the whole horizontal. The young shoots, which may stand more nearly vertical, reveal the true order. As this habit is retained in the Irish Yew the leaves are never distichous on that variety.

Floral Phyllotaxis.-As the parts of flowers are all identical in origin with leaves it is natural to suppose that they will follow the same laws. This they do, with certain modifications, in order to adapt them for their new functions. We have seen why 3's prevail in monocotyledons, and in dicotyledons 5's. The reason is that 5's are cycles of the ? arrangement and 3's of the. But, instead of each cycle, i.e. floral whorl, being exactly over the preceding, the "law of alternation now intervenes; so that each whorl alternates, i.e. in a "decussate" manner, with the next. Hence when a whorl happens to be suppressed this alternation appears to be violated, as stated in the case of the five stamens of a Primrose. If 2's or 4's occur in flowers of dicotyledons, they generally result from opposite leaves, as in Lilac.

By the way the petals are folded in the bud various modifications become revealed. If the reader draw a plan of the arrangement on a spiral containing two coils, and place the Nos. 1, 2, 3 at an angular distance of 144°, he will find that by a slight alteration the second petal (represented by a crescent) may have one edge underlying the third petal. Then, if the fourth similarly underlies the fifth, we get the socalled imbricate whorl, i.e. each petal overlaps the next. But when the first is also underneath the third then the convolute or "contorted" condition is reached, characteristic of mallows, flax, convolvulus, &c.

Various modifications are to be seen in papilionaceous flowers of the Pea family, and especially in flowers with coherent pètals, as the Snapdragon, &c., but for details of these I must refer the reader to my paper elsewhere.*

Elaborate calculations as to the number of parts of flowers in many different individuals of the same species have been made to ascertain their relative sequence. Thus on counting the rays or pedicles of umbels of Torilis Anthriscus the maxima were 5, 8, 10 (=2×5); while the number of ray petals of Chrysanthemum segetum gave 13, 21, 26 (=2×13), 34. In the case of the stigmatic rays of poppies 13 is the maximum. With the ray flowers of the ox-eye Daisy, 21 far outstripped the others. In the Cowslip, 3, 5, 8, 10(-2x 5), 13 were the maxima, 5 being the chief.

It will be at once seen that these maxima correspond with the different phyllotactical arrangements; but why, in fixing the number of parts in a whorl, Nature should endeavour to retain what may be called "cycular" numbers is unknown. The above experiments in calculations * "On the Origin of Floral Æstivations," Trans. Linn. Soc. vol. i. 1876, p. 177.

only illustrate what may be observed in nature. A composite with ray florets often has eight only. This was the case with the wild Dahlia and Zinnia, but under cultivation the number rapidly increased. Chrysocoma has regularly five only, thereby mimicking a five-petalled flower. One common by road-sides near Cape Town, and in a semi-starved condition, always bore three ray florets only. A correspondent informs me that in counting the ray florets of a South European marigold the maxima were 21 and 34, being correlated with their habitats. Thus 21 is the usual or typical number; but near the sea the first maximum rose to 26 (=2×13) and 34. "The ox-eye daisy has generally 21 florets at the level of Lake Como; but at a height of 400 to 500 feet on the mountain a form is developed during the height of the flowering season with 34 rays, reverting at the end of the season to 21. The number of ray florets in different composites varies with the environment."

With regard to the 3 sepals and 8 petals of Ranunculus Ficaria, the former arise from the fact that the plant is monocotyledonous, as stated above, but the change from 3 sepals to 8 petals is due to the fact that the angle of 120° (3) is nearer to 135° (3) than 144° (2), so that the change is easier from the to arrangement than to the .

Conclusion. I have now endeavoured to give a brief and as clear account as I can of the main facts and principles of Phyllotaxis. But, if we venture to search for a cause of such definite and exact arrangements of leaves, it will probably be fruitless, for as yet no satisfactory explanation has ever been given. It is not enough to say that it is a wise arrangement that leaves should not all be over one another, so as to exclude the light and air, and impede one another's functions; but why are there so many different ways of doing it, as represented by the fractions,, 3, &c.?

ARCHES, PILLARS, AND PERGOLAS.

By Mr. WALTER P. WRIGHT.

[Summary of a lecture delivered on July 11, 1907.]

THIS subject could be dealt with from the descriptive, the artistic, or the practical point of view. I might merely give particulars of existing examples, say what I think they ought to be, or describe how to construct and clothe them. But it would be of little real value to speak at length of the pretty arches or beautiful pergolas in various parts of the country, and I will pay most attention to practical matters.

First, however, let me say that the word "art" is used in far too loose a way in connection with flower gardening. Most people seem to think that they get an artistic garden if they simply leave out "geraniums,” and that when they have stuck in some arches, in however incongruous a way, their handiwork is complete. Often the arches are of wire; I prefer my ironmongery in the form of spades and hoes.

As an example of a modern artistic garden I recently saw a straight walk spanned with a series of disconnected wire arches, each with an anæmic and decrepit example of Crimson Rambler Rose. Why arches in such a place? The natural and appropriate use of arches is to mark divisions of gardens. The proper thing for a straight walk is a pergola. Arches out of place, of the wrong material, and with unhealthy plants upon them are ugly and incongruous.

When we have made up our mind that wooden structures are more suitable than metal ones we bring ourselves to practical questions. What is the best material? Where are we to look for it? What may we expect to have to pay for it? The best natural material is larch, because it is straight and is a recognised market commodity. Timber merchants in country districts often stock larch poles about 22 feet long, which, when cut into two, will make two uprights of suitable length. It is the truest economy to purchase a supply of long poles and cut them up. Do not let the vendor peel them; the bark is best left on, as the poles then have a more natural appearance. Larch is not a durable wood naturally, but if 3 feet at the bottom of the pole be peeled and treated with some preservative it would last for a very long time. Some might ask, Why treat 3 feet? Will not half that length suffice? Arches, pillars, and pergolas must be constructed properly, and in order to have them firm and secure from being swayed by the wind, which is very bad for the plants, the upright should be sunk into the ground nearly or quite 3 feet. With an 11-foot pole there will be 3 feet in the ground and 8 feet out of it, which is the orthodox height.

A common preservative is creosote. This should not be used for green wood, as the latter will absorb too much. Seasoned poles should be secured. The creosote should be used cold, as then it has no deleterious action on the tissues of the wood. It is better to soak the

E

ends of the poles in it for two or three weeks rather than to paint it on, and when treated thus the poles are well-nigh imperishable, but it is not innocent of causing root injury. A safer plan is to get ordinary coal tar, ready boiled, paint it on the bottoms of the poles, and dust with sand. This answers well, and is considerably cheaper than creosote, enough for a considerable number of poles being obtainable for 5s. A third and still more economical way is to char the ends of the poles in the garden fire.

The question now arises: Where shall we get the material? Obtaining poles is not so serious a business as is generally supposed. They can often be got at timber yards in the larger country towns where good nurseries exist. Or they may be obtained at timber sales, which are held in many districts in autumn. For a 15-foot to 18-foot larch pole, rough, one might pay 6d. at a sale, and for a 22-foot pole at the timber merchant's 2s. to 3s. Even at the higher price, viz. 3s., one cannot say it is dear, as two 11-foot poles are thus obtained for 1s. 6d. each. In districts where there are woods, dead spruce may often be obtained cheaply by getting into touch with the forester, and it can be made good use of. So far as durability is concerned oak is undoubtedly the best, but it is very expensive, 11-foot posts costing 3s. to 4s. each. Even an oak post needs the base dressing. When putting arches together one wants a certain amount of short, gnarled, crooked, rough material for the upper part. This is often difficult to get, but it can be obtained at many wood yards in districts where there are large gardens.

A special word should be said about pillars. They form a most valuable feature in flower gardens. Clothed with Clematises, as at Kew, they are exceedingly beautiful, and for giving variety they are splendid. An excellent plan is to have three good pillars in a rough triangle, 3 feet to 4 feet apart, of good, stout material.

Pergolas are annually becoming more popular. I might describe a pergola as a connected series of arches. It may be constructed in various ways, and one of the most solid and attractive that I know of may be briefly described. It is in the garden of the mayor of Canterbury, Mr. Bennett Goldney. On the top of each of the upright poles that form the supports for the pergola is laid a stout piece of unpeeled larch, about 18 inches long. On this meet the ends of three horizontal poles-two forming the continuation of the top side lines and the other connecting the two sides of the pergola across the walk. These three ends abut. On them lie the ends of two more poles-namely, the top diagonals. All these are securely spiked together, forming an absolutely taut, substantial, and homogeneous erection.

The ordinary pergola has the upright poles 8 feet out of the ground and 8 feet apart, but of course this can be varied considerably to suit individual requirements. The upright poles on each side of the path are connected by two lines of horizontal poles, and these in turn are connected with cross poles of lighter weight; 6-inch material for uprights and 3-inch or 4-inch for cross poles are generally used, but of course much more substantial structures are erected where cost is not the first consideration.

We now come to the important question of the most appropriate kind of plants for arches pillars, and pergolas. One cannot recommend the same

kind of plants for all. For arches the first essential is plants that produce long canes, Crimson Rambler' Rose being a good example. For pillars

we want something that throws out vigorous lateral growths, a typical plant being the old but beautiful rose Félicité-Perpétue.' For pergolas