When examined under the microscope, these cotyledons, the alimentary part of the seed, have the appearance represented in Fig. 6, where ep is the epidermic layer and cp constitutes the uniform parenchyma of the cotyledonary leaf.  This parenchymatous mass consists of oval cells filled with fatty matter and granules of aleurone.

According to some chemical researches made by Professor Schlagdenhauffen, the pulp has the following composition per 100 parts: 

    Fatty matter 2.407
    Glucose 33.92
    Inverted sugar 7.825
    Coloring matter and free acids 1.300
    Albuminous matter 5.240
    Gummy matter 19.109
    Cellulose 8.921
    Lignose 17.195
    Salts 4.080
                                   -------
    Total 100.000

The salient point of these analytical results is the enormous quantity of matter (nearly 60 per cent.) formed almost exclusively by sugar.  It is not surprising, from this that this product constitutes a food both agreeable and useful.

An analysis of the entire seed, made by the same chemist, has given the following results: 

Solid fatty matter 21.145
Unreduced sugar 6.183
Undetermined matters 5.510
Gummy " 10.272
Albuminoid " 24.626
Cellulosic " 5.752
Lignose and losses 20.978
Salts 5.534
-------
Total 100.000

The presence in these seeds of a large quantity of fatty matters and sugar, and especially of albuminoid matters (very nutritive), largely justifies the use made of them as a food.  The innate instinct of the savage peoples of Africa has thus anticipated the data of science.—­La Nature.

* * * * *

THE HEIGHT OF SUMMER CLOUDS.

A knowledge of the heights and movements of the clouds is of much interest to science, and of especial importance in the prediction of weather.  The subject has therefore received much attention during recent years from meteorologists, chiefly in this country and in Sweden.  In the last published report of the Meteorological Council for 1885-86 will be found an account of the steps taken by that body to obtain cloud photographs; and in the Meteorologische Zeitschrift for March last, M.M.  Ekholm and Hagstrom have published an interesting summary of the results of observations made at Upsala during the summers of 1884-85.  They determined the parallax of the clouds by angular measurements made from two stations at the extremities