ADDITIONAL NOTES. 214 expressed by writing (k[lh]l). This makes it difficult to distinguish between (tsh) and (t[jh]j) in such words as 4 nature/ § 244, p. 85. I have now, by the help of Professor Storm’s directions, acquired the 4 thick’ Norwegian l (§ 132, note b above). It is rj. finished off with mo¬ mentary contact of the tongue-tip and the inside of the arch, the tongue moving forwards all the while, and seems, therefore, to be a sort of inversion of the Japanese r. The second element is quite instantaneous in its formation, almost like a single strong trill. I would write this sound (rjflj.]) rather than (r4-[dj]), there being no stoppage of the mouth passage, but only contact of the tongue-tip. The effect is mainly that of (r) with a slight (1) quality. P. no, 1. 3. The vowel in 4 bird5 ought strictly to be represented by oe, which I have adopted in the Appendix. P. in, 1. 12. Add 4 rather the Norwegian open u! P. 123. Professor Storm entirely repudiates most of the values assigned by Bell to the French nasals, accepting those first given, which represent my own analysis also. He prefers (as I do now) (seq) to (æg), and also (0eq) to (æhq). P. 144. The Icelandic unaccented u in 4 mönnum/ 4 rikur ’ seems to be generally simple (eh) or (eh), and the i of 4 riki ’ seems to incline towards (e1). P. 146. (k) and (g) ought to have been written (kj) and (gj) consistently with § 149. P. 147. The notation (se'gi) is the correct one, for the on- as well as the off-glide seems to be really whispered. P. 153. In Swedish Broad Romic 4 daaga/ &c. might be written simply 4 daga,’ the length of the vowel being taken