The leading characteristics of the German style are the double apsidal arrangement of plan, the multiplication of small circular or octagonal towers, combined with polygonal domes, at the intersections of the transepts with the nave, and the extended use of galleries under the eaves of the roofs both of the apses and of the straight sides. The most ornamental parts are the doorways and the capitals of the columns. The latter surpass in beauty and in richness anything of their kind executed during the Middle Ages, and, though sometimes rude in execution, they equal in design any capitals ever invented. These only required the experience and refinement of another century of labour to qualify them to compete successfully with any parts of the pointed style of architecture which they borrowed from the French, and which in the course of time entirely superseded their own native style. CHAPTER II. BASILICAS. CONTENTS. Plan of St. Gall—Church at Mittelzell in island of Reichenau— Romain-Motier— Granson—Church at Gernrode—Trèves—Hildesheim—Cathedrals of Worms and Spires—Churches at Cologne—Other churches and chapels— Double churches—Swiss churches. ST. GALL. AS just mentioned, the history of Gothic architecture in Germany commences practically with Charlemagne; and, by a fortunate accident, we are able to begin our account of it by quoting from a contemporary illustration of the greatest interest and importance. In the library of the monastery of St. Gall, in Switzerland, a manuscript plan of a great monastic establishment was found by Mabillon in the 17th century, and published by him in the second volume of the ‘Annals of the Benedictine Order.’ The name of the author is not known; but, from some dedicatory verses on the back, it appears certain that it was sent to Gospertus, who was abbot of the monastery, in the beginning of the 9th century, and who in fact rebuilt the church and part of the monastic buildings between the years 820 and 830. Mabillon conjectures that the plan was prepared by Eginwald, the friend of Charlemagne, and who was also the director of his buildings. It is by no means improbable that this may have been the case, though it does not seem possible to prove it.