This illustrates only the average condition. On a calm day in the case of a beach facing south in the South of England, I have obtained exactly the same readings in July as at Valparaiso in January, 112° at surface, 80° four inches deep.

NOTE 71 (page 479)
 
On the Buoyancy of the Seeds or Seed-vessels of some Chilian Shore Plants

(1) Nolana, probably paradoxa. Common on the beaches of Southern Chile. The ripe drupes have a somewhat fleshy outer covering which they lose when lying on the sand, and present themselves then as dark-brown angular “stones,” often five to six millimetres across. Inside the outer hard covering of the stone is a layer of spongy tissue which gives it buoyancy; but since these coverings are wanting at the scars marking the basal insertion of the drupe, the embryo seems insufficiently protected against injury during flotation in sea-water; and the seed-vessel at first appears to be only fitted for conveyance by the currents over a limited tract of sea. However, in a preliminary experiment on seed-vessels that had been kept a few weeks, I found that 30 per cent. floated after three weeks in sea-water. Subsequently, after drying for a year, the seed-vessels were again tested in sea-water, nearly all of them floating after three months’ immersion. Two of them, removed after six weeks’ flotation, germinated healthily. These fruits are common in beach-drift between Corral and Valparaiso.

(2) Raphanus, near R. maritimus. Growing near beaches in South Chile, and not infrequently represented in the stranded beach-drift by the pods, which in my experiments floated seven to ten days in sea-water, after drying some weeks.

(3) Franseria. A species common on the beaches of Valparaiso and Talcahuano. Its prickly fruits, after being kept six weeks, floated only two to four days. They are well suited for transport in birds’ plumage.

NOTE 72 (page 483)
 
The Southern Limit of the Mangrove Formation in Ecuador.

... The southern limit of the mangrove formation on the west coast of South America is usually placed at 4° S. lat.; but it is probable that the vicinity of Tumbez in lat. 3° 30ʹ S. would be more correct. Baron von Eggers would place it rather further to the north-east, near the frontier of Ecuador and Peru in lat. 3° 20ʹ S. I spent eight days in the locality last named and saw no evidence of the beginning of the mangrove-formation.

NOTE 73 (page 495)
 
Additional Note on the Temperature of the Dry Coast of Ecuador between Puna Island and the Equator.

... Baron von Eggers gives the mean annual temperature for El Recreo, about half a degree south of the equator, at 75° F., which is near that of Rio de Janeiro in lat 23° S. on the east coast of the continent. Mr. F. P. Walker has kindly given me the results of temperature-observations covering a period of ten years, taken in the room for testing cables at Santa Elena Point (2° 10ʹ S.), usually about 6·30 a.m. The range of the monthly means was 71° F. (August) to 79·1° (March), and the mean for the year was 74·8°. In that locality a typical daily range would be 65° to 80°; and Mr. Walker believes that a minimum of 59° has been recorded.

NOTE 74 (page 495)
 
Observations on the Temperature of the Humboldt Current from Antofagasta Northward, between January and March, 1904 (Fahrenheit scale)

The observations were usually taken at the anchorages, but in some places, as at Ancon and Puerto Bolivar, they were taken from a boat outside the roadstead.

If we wish to ascertain how the Humboldt Current retains its cool temperature as it advances through the tropics to the equator, a glance at the following table will show that the surface-temperatures can aid us but slightly, since they do not vary in accordance with the latitude, a subject further discussed below. We can, however, obtain some valuable indications from the deeper temperatures. Let us take for instance the plane of 60°. Whilst south of Ancon (lat. 11° 45ʹ S.) it was rarely deeper than four fathoms, north of this latitude it descends rapidly, being probably about ten fathoms down at Salaverri and Eten and about twenty fathoms deep at Payta, in latitude 5° S., where the Humboldt Current leaves the coast. Within the Gulf of Guayaquil it is probable that the plane of 60° would descend to nearer thirty fathoms, the region being outside the influence of the current.

Some interesting facts are also elicited from the variation of the surface-temperatures. When we were coasting along at a distance of five or six miles from shore the readings were fairly constant from hour to hour varying only a degree or so. But nearer the land, for instance, about two or three miles away, the variation from hour to hour amounted to two or three degrees, whilst within the limits of the anchorages, a mile and less from the coast, the change from hour to hour amounted to three or four degrees. Nor was there any uniformity at the same hour over the surface of a roadstead. The temperature would often rise or fall a degree every few boat-lengths. Sometimes the inshore water was the coolest and sometimes it was the warmest. Thus at Iquique the inshore water was three degrees warmer than the water half a mile out, whilst at Mollendo, when the temperature one-third of a mile off the shore was 70°, it was 63° close to the rocky coast. The same thing was exhibited at Pisagua, where the surface-water two miles out at sea was 61°, whilst close inshore at the anchorage it was 58°. It was evident that there was a considerable intermingling of the warmer surface and the colder, deeper waters on the coasts of Chile and Peru. This was particularly noticeable on a rocky, steep-to coast, or where there was an uneven bottom. At some places, indeed, the warm upper layer did not exist, the cold water welling up all along the coast. This was especially the case between the 22nd and 19th parallels of latitude, a tract of coast in which lie Tocopilla, Iquique, and Pisagua, and probably the coolest part of the sea-border at this season of the year.

During a fortnight spent at Ancon (11° 45ʹ S.), between January 27 and February 10, I paid considerable attention to the local climatic conditions, and especially to the temperature of the inshore water. The daily range of the air-temperature was only five or six degrees, the average minimum and maximum being 71° and 75·9°, and the mean for the period 73·5°. The mean temperature of the surface-water at the head of the pier, from observations taken at about 7 a.m. and 4 p.m., was 68·6°, or five degrees cooler than the air, the mean temperature in the morning being 69·1° and in the afternoon 68°.

The Ancon climate at this period is full of oddities and abnormalities, and in this way typifies much of the coast of Peru. Thus, since the heat of the day is tempered by the cool south-westerly winds which die away in the evening and give place usually to warm, light, northerly and north-westerly breezes, there is, as above remarked, but a small difference between day and night temperatures. The coldest time of the twenty-four hours is not in the early morning but at sunset. The sea off the beach is, on the average, much cooler than the air, which is not a normal state of things; and again, the water is often two or three degrees colder in the evening than it is in the morning, which is very unusual. Though the sea-border is practically a desert for the greater part of the year and has no rain, it is frequently enveloped in drizzling fogs or “garuas.” Judged from a European standard, things go by contraries on the coast of Peru; and this is entirely the effect of the Humboldt Current.

The temperature of the inshore waters of Ancon Bay varied considerably during the twenty-four hours. During the day, with the prevailing southerly wind, the cool waters of the current had free access to the bay, and swept around its border in their course north; but in the night, when northerly breezes occurred, the cold waters of the current were pushed off the coast and their place taken by the warmer inshore waters from the north; and this sometimes continued for a day or two. When the current again got mastery and its clean, cool waters filled the bay, the temperature of the water dropped suddenly five or six degrees, and the bay was filled with fish. At such times men in boats leave the beach, and in a few minutes, with hand-nets and baskets, they obtain thousands of the small fry. Other men, fishing with lines from the pier-head, seem ill-contented unless they can catch fish of the size of small mackerel at the rate of one a minute.

There can be little doubt that on the coasts of Chile and Peru the instincts of fish often lead them astray, on account of the sudden changes of temperature arising from the conflict between the warmer waters of the open sea and the cooler waters of the current. From the preceding remarks it will be inferred that sometimes the current is pushed off the coast for a while and its place taken by the warm waters from the north. At other times it dives down, so to speak, and flows at a deeper level, and warmer waters prevail both out at sea and inshore. At other times again, and this must be most disconcerting to the fish, the cold current suddenly appearing at the coast predominates at the surface for days together, and we have stretches of coast which, although lying within tropical latitudes, are washed by waters having the temperature of the temperate zone. It is to such causes that we must attribute the reckless habits of fish on these coasts. They are known to throw themselves on the beaches in thousands, where by their decay they taint the air long afterwards. Mr. Anderson Smith in his recent book on Temperate Chile vividly describes what goes on on such occasions at the port of Valdivia. At times the scene must be indeed a strange one, since huge octopi are rolled up on the beaches in numbers, and are regarded by the indigenes as deliberately seeking their death. Whether they commit suicide or not, “their beaks that blacken the edge of the sea-wash in places” afford a melancholy proof that their instinct has blundered.

The Mode of Observation.—A thermometer made on the Sixe pattern which I used several years ago for taking the bottom-temperatures of rivers, was employed for the deeper temperatures, and at critical depths the observations were always repeated. This instrument was compared after each set of observations with an ordinary thermometer graduated on the stem, which was compared with my standard thermometer provided with a Kew certificate.... The observations in the Panama Roadstead have been added for the sake of contrast.

NOTE 75 (page 496)
 
On the Stranded Massive Corals apparently of the Genus Porites found on the Coast of Peru and North Chile, at Arica (18° 25ʹ S.), Callao (12° 3ʹ S.), and Ancon (11° 45ʹ S.)

At Arica they occurred on the beach only. At Callao they also extended inland on the low spit at Punta for about 100 yards. At Ancon they were found not only on the beach but also twenty or thirty paces inland on the low adjoining plains. Their size varied from three inches to three feet. They were all more or less rounded by wave action, and were extensively burrowed by boring molluscs. Whilst some on the beach still displayed the dried-up soft parts of the boring mollusc, others inland were falling to pieces and undergoing chemical change. There was nothing to indicate that the corals were recently alive; and at Ancon they appeared to have been torn off a rocky spit of andesite that had become exposed on the beach during a recent movement of emergence, of which there is other evidence on this coast. Further particulars are given on page 496.

NOTE 76 (page 429)
 
Stranded Pumice on English and Scandinavian Beaches

Sernander, in his description of the Atlantic drift of the Scandinavian coast, refers to the occurrence of a small amount of true pumice. I have found solitary fragments of acid pumice well rounded by wave-action at Croyde Bay on the north coast of Devonshire, at the mouth of Salcombe Harbour on the south coast of the same county, and at Maenporth, near Falmouth, in Cornwall. Steamer slag, in some cases rudely simulating pumice, is common on all the South of England beaches I have examined. It is also common on the Scandinavian coasts, though seemingly regarded by Helge Bäckström, who is quoted by Sernander, as derived from the factories on the east coast of England. (See on these subjects a paper by Helge Bäckström, “Über angeschwemmte Bimsteine und Schlacken der nordeuropäischen Küsten”; Bihang till K. Sv. V. A. Handl. Bd. 16. Afd. 3, 1890; also a letter in Nature, about 1886, by H. B. Guppy.)

This small tree has a very wide distribution in the tropics, ranging from East Africa and the Mascarene Islands through India, South-eastern Asia, Malaya, New Guinea, and the Solomon Islands to Fiji and Tahiti. It is a plant that grows in inland open woods as well as amongst the littoral trees on the beach; and it is always doubtful (in Malaya, Fiji, and Samoa) whether to regard it as a shore plant or as an inland plant, different authors varying on this point. In Vanua Levu I formed the opinion that it is only an intruder amongst the littoral vegetation. In accounting for its distribution we have to choose between man, the bird, and the current. Though it may sometimes be noticed in native plantations, as I observed in the Solomon Islands, the tree has no special use; and the Solomon Island natives themselves indicated to me that the parrots that fed on the fruits of the tree aided in distributing the plant. The buoyant behaviour of the seeds, which are freed by the dehiscence of the bladder-capsules on the tree, is not constant. Whilst in the case of the seeds of littoral trees in Fiji I found that 30 per cent. floated after ten weeks, Prof. Schimper ascertained in the case (seemingly) of Malayan seeds that they sank at once. The seed-structure connected with the buoyancy is, as shown on page 105, accidental in character, and reference is made on page 20 to other plants of doubtful littoral reputation, in which the buoyant qualities are variable. The occasional buoyancy of its seeds will only, as I think, explain its occasional station at the coast; and I agree with Prof. Schimper (p. 156) when he attributes its wide distribution to birds, the seeds being hard, crustaceous, and about three millimetres across.

NOTE 78 (page 436)
 
On the “Sea”: an Unidentified Wild Fruit-tree in Fiji

This is a fair-sized forest tree common in places in the lower forests. I have never been able to identify it; but a “putamen” which was sent to the Kew Museum was named Spondias with a query. It is to be hoped its true botanical name will be discovered by one of my successors. Seemann places it amongst the “desiderata” concerning which further information is needed. The fruit is a drupe 2 to 2¹⁄₂ inches long possessing a pleasant fruity odour and inclosing a hard two-celled stone about 1²⁄₃ inch long, one cell containing a large fleshy seed covered with tawny hair, the other filled with the hair only and containing no seed. The Fijians say that these fruits, large as they are, are swallowed by the fruit-pigeons, the stones being found in their gullet. The leaves are distichous, alternate, lanceolate, eight or nine inches long, glabrous and dark green above, and covered below with a whitish woolly matted tomentum. The empty stones are not uncommon in the stranded beach-drift.

A number of observers, beginning with Humboldt, in his Ansichten der Nature, and including Seemann, L. H. Grindon, Ridley, Beccari, and others, have referred to what is called “stenophyllism” in plants. These willow-leaved river-side plants are found all over the globe, such plants usually growing close to the water’s edge in situations where they are liable to be more or less submerged when the river is in flood. Seemann, Beccari, and Ridley mention more than two dozen genera belonging to a great variety of orders, and including Acalypha, Antidesma, Calophyllum, Eulalia, Eugenia, Fagræa, Ficus, Garcinia, Ixora, Lindenia, Melastoma, Podocarpus, Psychotria, &c., all tropical, and represented either in Fiji, Borneo, or in the Malay Peninsula; whilst my readers will recall amongst temperate floras river-side plants of the genera Epilobium, Lythrum, Salix, &c., possessing the same form of leaf and the same station. The genus Eugenia comes under this category in Fiji, Borneo, and the Malay Peninsula, with reference to one or more of the species. In Fiji, species belonging to the genera Lindenia and Dolicholobium especially attracted my attention in this respect. It is noteworthy that several of the Bornean plants and some of the Fijian plants here concerned are endemic. Just as I have remarked in the question of the buoyancy of seeds and fruits, that not all water-side plants have buoyant seeds or fruits, but that nearly all plants thus endowed are found at the water-side, so we may say of the willow-leaved plants, that not all river-side plants have the willow-form of leaf, but that plants thus characterised gather at the river-side. Beccari and Ridley regard this willow-form of leaf as the result of adaptation. Seemann remarks that we have here the old question whether the webbed feet of a duck are the cause or the effect of the bird’s swimming; and I take the same position. (See Seemann’s Flora Vitiensis; Ridley in Trans. Linn. Soc. Bot., vol. iii. 1888-94; and Beccari’s Nelle Foreste di Borneo, 1902, or the English edition of 1904.)

NOTE 80 (pages 255, 504)
 
Mr. Perkins on the Hawaiian Lobeliaceæ (Fauna hawaiiensis, vol. I.)

My view, that the early Hawaiian Lobeliaceæ acquired the monstrous form of their flowers in the humid forests of a later age, is supported by the observations of Mr. Perkins on the connection between the highly-specialised nectar-eating Drepanids of Hawaii and the highly-specialised flowers of the Tree-Lobelias, a subject further discussed in Chapter XXXIII. This naturalist ascertained, in the case of one of the trees, that fertilisation could only be effected by these birds. So close is the biological connection between the Drepanid and the Tree-Lobelia, that Mr. Perkins finds here in part the cause of the development of the most remarkable forms of the birds. The botanist, also, would not dissociate the plants from this conclusion. There would be every reason to look for abnormal growth in birds and plants when the bird depends on the flower for its food, and the flower is dependent on the bird for its pollenisation. It is through such guises that the zoologist and the botanist have to penetrate when establishing the systematic affinity.

NOTE 81
 
On the Vertical Range of some of the most Typical and most Conspicuous of the Plants in the Forests on the Hamakua Slopes of Mauna Kea, Hawaii

During a descent of this mountain on its north side to near Ookala, the conditions were unusually favourable for recording the range of altitude for some of the plants easily recognisable.

Acacia koa began at 6,700 feet, and extended down to 2,300 feet.

Rubus (“akala”) began at 6,500 feet, and extended down to 2,500 feet.

Cheirodendron (“olapa”) began at 6,400 feet, and extended down to 2,200 feet.

Cyanea, a lobeliad growing on trunks of tree-ferns, began at 4,000 feet, and extended down to 2,300 feet.

Freycinetia began at 3,850 feet, and extended down to 2,000 feet.

Asplenium nidus began at 2,800 feet, and extended down to 2,200 feet.

Aleurites moluccana began at 1,800 feet, and extended down to 50 feet.

Metrosideros polymorpha, ranging through all the zones.

NOTE 82 (page 416)
 
Aboriginal Weeds^[7]

NOTES 83-89 omitted

NOTE 90 (page 29)
 
On the Buoyancy of the Seeds of Euphorbia amygdaloides and E. segetalis

The seeds of both species have no proper buoyancy, and display no structure in their testas suggesting it; though, through the shrinking of the nucleus, a temporary floating power may be acquired with less mature or imperfect seeds. They support the general principle indicated for the British species on page 29.

NOTE 91
 
Mr. E. Kay Robinson on the Dispersal of Aster tripolium

According to this naturalist, the seeds of this plant are eaten in winter by snow-buntings on the English east coast. In reply to my query he tells me that the “draggled fluff still containing seeds” might easily adhere to birds (The Country-Side, Sept. 30, 1905).