APPENDIX 555 3rd waterfall is called Gelandri (reXavSpi), which in Slavonic means noise of the fall. Only one name is given, and it is said to be Slavonic. But it obvi- ously represents the Norse participle gellandi, " the echoing " ; so that the Slavonic name (probably nearly the same as the modem name zroneti with the same meaning) is omitted. Constantine's usual formula is 'VaxriffTi nkv . . . iKAa^auicTTl 5e ; but in this place he changes it : rhi' Keyouei/ov TfKavSpi, h ipjx-qvivirai 'S.KKa^iviffrl fixos (ppayixov. I would suggest that 0ivirs or (TBi-vns or something of the kind ifell out after '2Ka$ivi(rTi 4th waterfall is Aeifor ('Aei<p6p, so in Paris Ms. 2009) in Russian, and Neasit iHeaff-fiT) in Slavonic, — so called. Constantino says, because pelicans make their nests in the stones. The Old Slavonic for pelican closely resembles tfeaartT, but the fall cannot have been called pelican ; this must have been a misinterpretation. Thomsen very ingeniousl}- suggests that the true name corresponded to the modern Nenasytets and meant insatiable (a name appropriate to the nature of this rapid) ; while Aeifor (ei-forr) meant ever- forward, ever-precipitate. 5th waterfall is Varuforos [^apov<)6pos) in Russian, Yulne prach (^ovvr]irpax) in Slavonic; "because it forms a great lake," or, if we read SiVtjv for KifjLifqy, "because it forms a great vortex ". Both words can be recognised at once as meaning " wave-fall . 6th waterfall is Leanti (Aedyri) in Russian, Verutze {Bepovr^ri) in Slavonic, meaning "the seething of water" i^pdcrua vepov). Verutze is obviously from v'rieti, to boil. Thomsen explains Leanti as the participle hlaejandi, laughing. In this case the meaning of the two names is not identical. 7th waterfall is Strukun (IrpovKow, so in Paris Ms. 2009) in Russian, Xapreze itiaTrpeQi]) in Slavonic, meaning a small waterfall. Thomsen identifies Strukun with Norse strok, Swedish st-ruk, a rapid current (especially where narrow — as in the case of this rapid) ; and suggests that the Slavonic name might be connected with brz, quick. I suspect that (No-) TTpf^ij represents a diminutive of porog, prog (waterfall). 16. THE ASSISES OF JERUSALEM— (P. 317) It is agreed by most competent critics of the present century that Godfrey of Bouillon neither drew up the Assises of Jerusalem as they have come down to us nor put into writing any code of law whatever. This is the opinion of such special students of the Crusades as "VVilken, Sybel, Stubbs, Kugler, and Prutz ; and recently it has been very forcibly put by M. Gaston Dodu in his Histoire des Institutions monarchiques dans le royaume Latin de .Jerusalem 1099-1291 (1894). In the first place, we find no mention of such a code in contemporary sources; the earliest authorities who mention it are Ibelin and PhiUp of Novara in the 13th century. Then, supposing such a code had been compiled, it is hard to understand why it should have been placed in the Holy Sepulchre and why the presence of nine persons should have been necessary to consult it. For the pur- pose of a code is that it should be referred to without difficulty. Thirdly-, the remark of William of Tyre as to the experience of Baldwin III. in judicial matters makes distinctly against the existence of a code. He says : juris con- suetudinarii, quo regnum regebatur Orientale, plenam habeus experientiam : ita ut in rebus dubiis etiam seniores regni principes eius consulerent experientiam et consulti pectoris eruditionem mirarentur (xvi. 2, cp. on Amalric i. xix. 2). The expression "the customary law by which the kingdom was governed" suggests that no code existed. Fourthly, if the code existed, what became of it? Ibelin and Philip of Novara say that it wats lost when Jerusalem was taken by Saladin in 1187. But the circumstances of that capture are inconsistent with the probability of such a loss. There were no military excesses and Saladin allowed the inhabitants a delay of forty days to sell or save their property before he entered the city (Emoul, c. 18 ; cp. Dodu, p. 45). It is highly unlikely that the Christians would have failed to
