mena, indulging his fancy in forming conjectures about their nature and origin. This circumstance, he himself states, first inspired him with a desire to study meteorology; and we can perceive in these solitary meditations, one of the causes which tended to give their fanciful complexion to many of his subsequent speculations.
After continuing his physical studies with much ardour for several years, he at length appeared in the character of an author. His "French Flora, or a brief Description of all the Plants which grow naturally in France," was published in 1778. The immediate occasion of this work was a desire to furnish his fellow-students with a system of arrangement which should lead with greater ease and certainty to the determination of plants than any then in use. For this purpose he adopted a modification of the binary or dichotomous method, the principle of which consists in arranging natural objects by their positive and negative characters, dividing and subdividing always by two, and allowing a choice only between two opposite characters. Although this plan is, of course, highly artificial, and ill calculated to throw light on the affinities and analogies of objects, yet it is much recommended by its extreme simplicity, which adapts it to the comprehension of those who have but little acquaintance with the technical and descriptive language of natural history. If judiciously applied, it affords an easy index to particular genera and species, and renders the subject at once accessible without any
