dyer's standpoint, in yielding little or no hematoxylin, but, instead, a yellowish-green pigment which is of no value and which, when admixed with the commercial extract, reduces the characteristic tinctorial properties of the latter. Chips of the "bastard" logwood present a yellow, pale pink, white or even chocolate-coloured surface instead of the dark red or deep purple, bronze-tinted colour of the best Jamaican or Mexican logwoods of commerce. There appears to be considerable uncertainty, even when the trees are cut down, as to whether a tree is really a "mulatto" ("bastard") tree or not. What is known as a "mulatto" tree is frequently dark enough when first cut to lead one to believe that it is a good redwood tree, but instead of darkening with age as all the good wood does, it remains the same colour or becomes lighter rather than darker. The "bastard" tree seems to be perfectly dry, and even when the chips are soaked for a long time in water, they give out no dye.*

Various theories have been advanced to explain the apparent increase in the "bastard" logwood in Jamaica. Professor F. S. Earle, after a thorough study of the situation in Jamaica, came to the following conclusions :†

I. "Logwood is a variable plant showing marked differences in form, colour and texture of leaf; time of blooming; form and extent of ribs on the trunk; colour of bark and especially in the colour and dye-producing quality of the heart-wood. Four well-marked varieties are said to be recognized in Honduras and three are usually recognized in Jamaica, but there are many other intermediate forms." 2. "Bastard" wood is not the result of disease or of any lack of vigour. The trees producing it are perfectly healthy and normal. 3. "It is not the result of soil or climatic conditions, since 'bastard' and normal trees are found growing side by side under absolutely identical conditions."

4. "It is not the result of immaturity. Aged trees may produce 'bastard' wood, while in normal trees the heart-wood, as soon as formed, contains a good percentage of hematoxylin. These facts seem to point to heredity as the probable cause of the trouble. That is, that certain trees produce only 'bastard' wood because they grow from the seed of a 'bastard' tree; or in other words that 'bastard' logwood represents a variety of Haematoxylon campechianum that normally produces little or no hematoxylin, just as one Honduras variety has smaller, shorter, thinner and lighter coloured leaves."

Some time before Professor Earle made his investigations in Jamaica we began, at Dr. MacDougal's suggestion, a comparative study of logwoods from that island, in the hope of finding definite chemical differences, other than purely tinctorial ones, between "red logwood" and the "bastard" variety. Unfortunately our

*Cradwick Report to the Chairman of the Experiment Station. Kingston, Jamaica 1902 (April 4.)

Earle: Journal of the New York Botanical Garden, 4: 3, 1903; reprinted in Bulletin of the Department of Agriculture, Jamaica, 1: 30. 1903.

work in collaboration was soon unavoidably interrupted. We present here very briefly, however, such of our notes in this connection as may be of general interest.

ELEMENTARY COMPOSITION OF HEART-WOOD.—Elementary analysis of typical samples of (1) the red logwood of commerce (2) a "bastard" variety somewhat resembling it and (3) a second specimen of the "bastard” type yielding hardly any pigment to water gave the following results :

:

TABLE I.

PERCENTAGE ELEMENTARY COMPOSITION OF SUBSTANCE DRIED TO CONSTANT WEIGHT AT 110°C.*

The most significant feature of these results is the decreasing amount of carbon in the "bastard" wood. The differences are too slight to warrant any emphasis, but are such as might be due to a lower percentage of hematoxylin, which is a pigment of high carbon (and low oxygen) content-C,, H1, O.

14

The data of the second series of analysis, given in TABLE II, show that the wood was not decomposed in the process of drying to constant weight at 110° C. (first series) and that, therefore, the previous results were not influenced by that procedure.

GENERAL COMPOSITION OF SEEDLINGS.-In TABLE III we present the results of some analyses of seedlings of "red" logwood

*Only heart-wood was employed in this work. This was converted into sawdust and only such portions as passed through a very fine sieve were taken for analysis. The methods of analysis were those which are now in general use.

†The figures for carbon and hydrogen are calculated (from the data of direct analysis), for ash-free substance.

Calculated, by difference, for ash-free substance.

and of the "bastard" variety. The condition of the seedlings at the time of analysis is shown in FIGURE I. The outward appearance of the two kinds of seedlings was practically the same. Likewise, the differences among the figures in our table for general chemical composition are too slight to warrant any other conclusion than that the seedling metabolism was, in general, essentially the same in both varieties. The analyses were made 12 months after seeds were planted.

*Analyses were made by the usual drying and incineration methods. The portions subjected to comparative analysis were approximately of the same morphological location in each variety. The most significant differences seem to be the slightly larger proportion of water in the "red" wood and the relatively greater quantity of solids, especially inorganic matter, in the "bastard" samples.

Bastard.

CONCLUSIONS FROM THE GENERAL ANALYTIC DATA-All of the preceding analytic results make it evident that the chemical differences existing among these logwoods are quantitatively very slight. They also make it appear probable that the variations in the different samples of the wood are chiefly variations in the

FIGURE I. Seedlings of logwood, one year old. a. "red logwood." b, "bastard logwood." Both grown from seeds obtained from "Old Hope" p antation, Jamaica. chemical characteristics of the pigments themselves, which as is well known, possess as a rule high tinctorial qualities even when they occur in only very small amounts. Our results in this connection would also indicate that there are no striking structural differences among these varieties of logwood. They suggest, likewise, that even metabolic tendencies in these logwoods are essentially the same, varying only, perhaps, in the course of events which involve relatively slight quantities of pigment.*

TINCTORIAL DIFFERENCES. The foregoing results having shown that the differences among these logwoods were chiefly if

These conclusions are in harmony with those drawn from other standpoints by Professor Earle (quoted on page 242). They were arrived at independently by us and were included in our report, in December, 1902, to the Botanical Society of America before we were aware of Professor Earle's deductions. Science, II. 17: 338. 1903.

not solely tinctorial, we next endeavour to ascertain the extent of the pigmentary variations.

Our first experiments in this connection were efforts to determine the relative tinctorial intensity of extracts of different samples of heart-wood sawdust made with equal volumes of various solvents under similar conditions of temperature, shaking, etc., from the same quantities of material dried to constant weight at II0° C.* Among the samples were several inferior qualities of red wood from dead and decaying trees.

TABLE IV gives our first results in this connection. The figures in that table denote the relative positions in a series of ten extracts I indicating weakest colouration, 2 the pigmentation of next higher intensity and so on to 10 showing the most decided tinctorial effect.

The shade of colour varied with each extractant, as would be expected. The following observations were made in this connection, on the colour of the series of extracts referred to in TABLE IV. I. Water-slight yellowish-brown to deep reddish-brown.† II. 0.2 per cent. HCl-faint yellow to orange.

III. 2.0 per cent. HCl-faint yellow through reddish brown to bright red.
IV. 0.01 per cent. KOH-chocolate colouration throughout.

V. 0.15 per cent. KOH-deep chocolate colouration throughout.

VI. 0.5 per cent. Na, Co, -chocolate colouration throughout; less than in
V, greater than in LV.

VII. Saturated borax solution-faint yellow to deep reddish-yellow.
VIII. Ether- faint yellow to orange.

IX. Absolute alcohol-faint yellow to red.

X. Acetone-faint yellow through greenish-yellow to yellowish-red.
XI. Acetic ether-faint yellow to deep reddish-yellow; brighter than in VII.
XII. Chloroform-no colour in some, faint yellow in others.

XIII. Benzol—no colour in any.

TABLE IV.

RELATIVE PIGMENTATION OF VARIOUS KINDS OF LOGWOOD.

Extractant.

1. Water

II. 0.2% HC1I.

III. 2.0% HCI.

IV. 0.01% KOH.

V. 0.15% KOH.
VI. 0.5% Na2 CO3.

VIII. Ether.

IX. Absolute alcohol
X. Acetone.

XI. Acetic ether.

Average

1.2 2 3 3 4 4.6 5.4 6 1 6.7 7.6 8.2 9.5

A—" Bastard” (very poor). B -"Bastard" (very poor). C-Immature wood of varying tints. D "Purple" (from tree on extremely poor marly bank; tree

*Drying occurred rapidly and seemed to have no transforming effect on the dust. This fact was noted before in another connection (page 243).

†The colouration intensities are indicated progressively from 1 to 10 (see TABLE IV) Individual exceptions are not referred to.

This sampl contained several pigments. One of these was purplish and quite unlike any in the other samples. The pigment was especially soluble in water. It was not ordinary hematoxylin.