NOTE.--In view of the late appearance of the Adibuddha (probably the tenth century), I have thought it desirable to state that the theory of the general design of the Boro-Boedoer contained in the text is based upon a very interesting conversation which I had with M. Groeneveldt, who is a member of the Council of Netherlands India and Director of the Museum at Batavia. Professor Rhys Davids has pointed out an interesting distinction between the Boro-Boedoer and the Buddhist shrines in India, viz. that, whereas the cupolas at Boro-Boedoer are hollow, the _dagabas_ of British India are always solid. In the Annex will be found a detailed account of the various routes and the cost, etc., of travelling from Batavia to the temple districts in the centre and east of Java. ANNEX TO CHAPTER V. THE ROUTES TO THE TEMPLES. Supposing that the traveller has been landed at Batavia, and wishes to visit the ruins in the east of the island, he will have the choice of three routes. First, he may sail by a Netherlands India boat to Samarang (or Soerabaia, if, as often happens from December to February, it is impossible to land at the former place owing to the surf); this occupies about thirty-six hours. There is an excellent hotel at Samarang--the Pavilion--where the night can be spent, and the following day the train will carry him to Amberawa, a distance of 50 miles by rail (or 30 by road). Here the railway stops, and a carriage must be taken to Magalang, the next town (with splendid views of the two volcanoes, Merbaboe and Merapi), which is some 20 miles further on, and where a halt must be made for the night. Ten miles' driving will take him to the Boro-Boedoer; the drive is one of extraordinary beauty. After visiting the Boro-Boedoer and the neighbouring temple of Mendoet, it is usual to return by way of Djokja (25 miles), which is the centre of numerous ruins. If, however, it is intended to travel overland, there are two routes available. The first is the regular posting route along the northern coast; the second lies to the south, and is perhaps more interesting. If the regular route is chosen, the traveller will proceed by rail as far as Bandong, a distance of some 90 miles; and then drive to Cheribon (80 miles), a place on the northern coast; and then, following the coast-line, from Cheribon to Tegal (40 miles); from Tegal to Pekalongan (35 miles); and from Pekalongan to Samarang (68 miles). In all these places there are good hotels, but two horses, and in some places four (as in the last stage, where the road passes over mountains), would be necessary. Such a journey in a carriage would cost (apart from hotel expenses) £20, or, if it were done in a cart (sadoe) and two horses, half that sum.