ii6 



SCIENCE. 



[Vol. XXI. No. 526 



Beinet at that time students in the Gra<3uate Department of 

 Harvard University, we laid under tribute the varied and exten- 

 sive resources of the Harvard Botanic Gardens. 



We first entered upon the histological study of the petiole and 

 its groove, but soondis30vered that that pointof view alone would 

 yield but meagre results: that we must depend largely upon 

 actual experiment with plants showing this characteristic, and 

 even more largely yet upon careful observation of the plants 

 themselves in their habit of growth and mode of branching and 

 the arrangement of leaves and roots on the particular plant under 

 observation. 



In order to get a starting-point it became necessary to make 

 some guesses or suppositions as to the origin or purpose. Some 

 of these suppositions came after a time to be so strengthened as 

 to justify us in calling them theories, a few of which 1 will give 

 with some of the facts supporting them. 



In looking for the origin we do not find the groove developed 

 as a characteristic structure earlier than in the Endogens. It is 

 true that from the earliest differentiation of tissue into leaf we 

 find in some instances the base of the leaf flattened and a strong 

 suggestion of a groove as seen in some mosses, lycopods, ferns, 

 and the bracts of the horsetails. 



Finding the groove well developed in the Endogens almost 

 without exception, and much less so in Esogens we are justified 

 in concluding that in the Endogens there exists a necessity for 

 such a leaf and petiole which is not found elsewhere. If, then, 

 we can discover what this necessity is we shall, at the same time, 

 have arrived at the origin of the peculiarity, for a plant by selec- 

 tion develops those structures best suited to its growth and per- 

 petuation. 



Without troubling myself at present about the reason for the 

 difference in the habit of growth of Exogens and Endogens, but 

 accepting these as we find them, keeping in mind always that a 

 change in the habit of growth of one part of a plant, due to a 

 change in surroundings, may necessitate a change in other parts 

 as well, we will notice the habit of growth of Endogens. 



Here we find plants with an unbranched stem, sometimes 

 quite long, but usually short and often reduced to a minimum. 

 The leaves are, as compared with the leaves of Exogens, few in 

 number but quite large, often extraordinarily so. In these facts 

 we can see some reasons for the grooved petiole. Like all leaves 

 they must be attached in some way, and nature will find the 

 most convenient way if at the same time her other purposes can 

 be subserved. Now, where space is at a premium as in very 

 short stemmed species or as in palms where the leaves are crowded 

 into a terminal bud, what better arrangement than grooved 

 flattened petioles overlapping each other can be suggested? 



But it is not only the most convenient way, it is also the strongest 

 way in which these could be attached. In a large number of 

 plants we find the base of the petioles so closely overlapping that 

 all are hound into so compact a mass that it would be almost im- 

 possible to pull out one without destroying the whole whorl or 

 I may say, the whole plant. This sheathing and overlapping of 

 the petioles is coincident with the groove, or, I had better say, the 

 groove is largely coincident with a sheathing base of the leaf. 



So then I would say that convenience of attachment for the 

 leaf and great strength for the plant as a whole are the fiist gains 

 to be noted. As examples I may cite most palms and other Endo- 

 gens with a very short stem and large leaves, or, to specify a few, 

 Latania Borbonica, Pandanus utilis, Homalonema caerulescens, 

 several of the Bilbergias, Crinum asiaticum, Pilcairnia hjslrix, 

 Bromelia pinguin, etc. 



Before leaving this subject of strength the gain to the indi- 

 vidual leaves by a sheathing base or grooved petiole needs to be 

 noted. 



We are all aware that from a given amount of material a 

 stronger structure is produced if arranged in the form of a hol- 

 low cylinder than if in a solid column. Nature makes use of the 

 hollow cylinder in grass stalks where strength and lightness is 

 desirable, why not then, so far as possible, use the same economy 

 in supporting those immense endogenous leaves, many of which 

 are subjected to tremendous strain because of the long petioles 

 necessary to carry them out into the sunlight? 



As a matter of fact vve do find them in all degrees from a com- 

 pletely sheathing base to half-cylinders, even the latter of which 

 is greatly stronger than the same amount of material in a solid 

 column. 



Then, again, that mode of attaching the leaf gives grenter 

 strength because one side of the petiole braces the other. That 

 could best be shown practically, yet one can easily conceive that 

 where, for instance, the extreme margins of the petiole are at- 

 tached to the stem 180° apart, each margin furnished with its 

 fibro-vascular bundles continuous into the stem, these marginal 

 fibro-vascular bundles act as guy ropes upon the long peti led 

 leaf as it is swayed by the breeze. 



We may further note the convenience of attachment with re- 

 spect to the fibro-vascular bundles themselves. These being con- 

 tinuous from the leaf into the stem and the leaf attaching in a 

 semi circle at a uniform height on the stem there is no necessity 

 for a convergence of these bundles at a given point. But a 

 greater number of bundles can pass naturally and directly into 

 that part of the stem down wBich they continue. As examples 

 illustrating this 1 would call your attention to corn and the 

 grasses generally. 



The groove having been formed in the base of the petiole it 

 naturally peBsists throughout its whole length and even in the 

 midrib of the leaf. Development in this manner, rather than a 

 change from a grooved to a cylindrical form, represents develop- 

 ment along the line of least resistance. 



It has been suggested that the groove represents the persistance 

 of a former condition or type and that at present the groove is 

 merely incidental and no longer functional. This does not seem 

 probable, however for as noted before we find little trace in 

 forms earlier than Endogens and again passing away in Exogens. 

 There is nothing to show that the groove is not to-day at its 

 highest state of development and differentiation. 



It is developed, as I believe, because of the unbranched and 

 often much shortened stem with the accompanying large leaves 

 found in the Endogens. In Exogens, where found, it seems to 

 perform the same functions already pointed out for Endogens, or 

 oftener it is here merely a persistance of a former habit of 

 growth. 



In the preceding statements I believe I have pointed out the 

 ancestral significance of the groove in the petiole. It now re- 

 mains to see whether, because of a change in surroundings, this 

 groove has secondarily taken on any new functions. 



It has been suggested : 1. That the groove and axillary pockets 

 thus foimed may be regions where the absorption of water takes 

 place. 2. That the groove guides the water to the young nascent 

 buds in the axils and that these may absorb. 3. That the groove 

 directs the water towards the main axis of the plant, and that in 

 such plants the root-hair area will be found near the main axis. 



If the above functions are found now to exist in some plants, I 

 think they have been acquired secondarily and comparatively 

 recently. The groove I consider coincident to and co existent 

 with the endogenous type of vegetation. Furthermore, at the 

 time when the endogenous type was the prevailing vegetation, 

 there was no necessity for the assumption by the petioles of the 

 above-mentioned functions. There are reasons for believing that 

 the regions where this type of vegetation took its rise were ex- 

 ceedingly well watered, and tne ground, being wholly shaded by 

 the denseness of the foliage, was at all points of nearly uniform 

 moisture, usually nearly saturated. In fact, this is still the con- 

 dition where this type of vegetation is most luxuriant. 



Careful microscopic examination of the tissues and their ar- 

 rangement in the groove and in the axillary pocket formed by 

 the petiole showed essentially the same structure as on other parts 

 of the petiole — cutinized always, sometimes as heavily as out- 

 side, usually without stomata in the pocket and few in the groove. 

 A large number were examined, and it seems justifiable to con- 

 clude that, ordinarily, the groove is not an absorbing region. 



There are, however, a few anomalous conditions and structures, 

 the use of which is difficult to comprehend. In the Bilbergias 

 we find the base of the petiole bearingalarge number of radiately 

 branched trichomes situated in small depressions in the epidermis. 

 It should be said that these petioles are so arranged on the stem as 



