The first memoir was on the Calamites, and controversy at once broke out. Williamson was from the first impressed by the manifest occurrence of exogenous, or, as we should now call it, secondary growth, both in the Calamites and the Lepidodendreae, groups which he was convinced were cryptogamic. The controversy with the great French school, headed by the illustrious Brongniart, is well known. As Williamson put it: "The fight was always the same; was Brongniart right or wrong when he uttered his dogma, that if the stem of a fossil plant contained a secondary growth of wood, the product of a cambium layer, it could not possibly belong to the cryptogamic division of the vegetable kingdom?[1]"
In England, however, the dispute was on different lines. "In August of 1871," says Williamson, "the British Association met at Edinburgh. At that meeting I brought forward the subject of cambiums and secondary woods in Cryptogams, with the result that my views were rejected by every botanist in the room." There followed a controversy in the pages of Nature, which is of some interest, as showing the state of opinion in England at that time. Williamson tells us in his autobiography the principle by which he was guided in his work: "I determined not to look at the writings of any other observer until I had studied every specimen in my cabinet, and arrived at my own conclusions as to what they taught." In spite of this excellent rule it is probable that he was at first unconsciously influenced by the views of Brongniart, which may have led him to attach too much systematic importance to the occurrence of secondary growth. At any rate he proposed at the Edinburgh meeting "to separate the vascular Cryptogams into two groups, the one comprehending Equisetaceae, Lycopodiaceae and Isoetaceae, to be termed the Cryptogamiae Exogenae, linking the Cryptogams with the true exogens through the Cycads; the other called the Cryptogamiae Endogenae, to comprehend the Ferns, which will unite the Cryptogams with the Endogens through the Palmaceae[2]."
It is curious to note in passing that his main divisions, so far as vascular Cryptogams are concerned, correspond to the
