^{page 131 note 1} In his Address to the first meeting of the Aryan Section of the Oriental Congress of 1874 he said, “More than half the words used by the Veddahs are, like Sinhalese itself, mere corruption of Sanskrit.” The correct spelling of the word Veddah is Væddsâ (æ pronounced exactly like our a in cat). Now in composition we find not væddâ but vædi, e.g. vædiminissu, “Veddah people,” and this form vædi is the original one, væsddâ being one of the curious nominatives with a doubled consonant which I spoke of at p. 12 of my first Note. The termination â I shall speak of further on; it is a late grammatical outgrowth, used to distinguish animate from inanimate nouns. The æ of vædi points to an original a or â, and this brings us to vyâdhaḥ, (not veddhâ which would give veda). The name Væddâ, therefore strictly means “archer,” or “sportsman.” Mr. Hartshorne informs me that Væddâ is not only the name given to these savages by the Sinhalese, but is also the name by which they call themselves,—a strong additional proof that the Veddahs are what I have always held them to be, wild Sinhalese.

^{page 133 note 1} In Pali we have Dâmiḷa=Sansk. Drâviḍa: bhamu, “eyebrow,” is perhaps bhram+u, but may it not be for a form bhrayu? vîmamsa is an instance of dissimilation, not of phonetic change, but the choice of v rather than of any other consonant is significant. Mr. Grose, in his article on the Etymology of Local Names in Northern India (Journal As. Soc. Bengal, part i. 1874), says, “dhîmar is the recognized literary form of the Sanskrit dhîvara, and at the present day villagers generally write Bhamâni for Bhawâni, though the latter form only is admitted in printed books,”

^{page 136 note 1} In the Kandyan provinces magê is pronounced by the lower classes mayi, the g being softened to y, and the ê to i.

^{page 136 note 2} I have accented the Sanskrit e and o in order that the reader may not forget that the unaccented e and o in Sinhalese are short.

^{page 136 note 3} I look hopefully to Mr. Beames to explain api.

^{page 142 note 1} It is worth mentioning here that in speaking of Buddhist priests namak ia used instead of denek, e.g. unnânsêlâ tun namak, lit. “priests three names.” This use of namak is honorific, and originates in the circumstance that when a man joins the priesthood he changes his name, taking a religious name, such as Dhammârâma “Delighting in the Law.”

^{page 142 note 2} Dedenek is generally compressed to dennek.

^{page 142 note 3} I think the most remarkable I know of is an English one. Such differentiations as thorough and through, condúct and cónduct, are commonplaces; but I do not think any one has pointed out that when “used” means “employed,” it is pronounced yûzd, while when it means “accustomed,” it is pronounced yûst.

^{page 147 note 1} Siñchana, sisana, sisina, hihina, ihina.

^{page 148 note 1} From the false analogy of forms like cîyate, kshîyate, dhîyate,. where the long î is correct.

^{page 151 note 1} With regard to the dative termination, Mr. Davids suggests its having originated in atthâya (arthâya). This accounts for the ṭa, but by no means for the h of haṭa, which cannot possibly be ignored.

^{page 151 note 1} As a specimen of the way in which Sinhalese philology has hitherto heen treated I will quote a passage from D'Alwis' Essay on the Origin of the Sinhalese Language (Journal Ceylon Branch R.A.S. 1867, p. 71):—“The Past Tense. Here I may allude to what Dr. Stevenson has pointed out, under this head, as an analogy which pervades all the Dravidian and the North Indian dialects, viz. that the past tense of the verb is marked by affixes and not prefixes, as in the Sanskrit. The Sinhalese is not without a prefix to form the past tense, as yami “I go,” and gi-ye-mi “I went” (!!); but I admit that the Sinhalese verb generally accords with the Dravidian in the peculiarity noticed by Dr. Stevenson. This admission, however, does not militate, etc. etc.” Giyemi is giya with a termination emi, which is the Sanskrit -âmi of the first person singular of the present tense, added on the false analogy of yami, keremi, etc., and belongs to the root gam, and not yâ. Giya is regularly corrupted from gata, compare siya = çata, liya= latâ, etc.

^{page 152 note 1} We have seen in this Note three cases of the vowel i in a termination carrying umlaut with it: 1st, the feminine termination î; 2nd, the passive termination in -îyati; and now the termination -ita of the participle of the perfect passive. It is interesting to compare â “came” = âgata with kæ “ate” = khâdita. In each case (as often happens when a long vowel is exceptionally retained) two final syllables are dropped, but only the syllables containing an i leave umlaut behind them.

^{page 152 note 2} Vyaghra is a tatsama, the Elu form is vaga. Lôkaya is also a tatsama.

^{page 154 note 1} Sahôdara and vâsin are Sanskrit tatsamas.

^{page 154 note 2} Mama laggâ, from my presence, lit. from my adhesion or union. Langin is Sanskrit lagnena, the Sansk. instr. term, having (as I have shown) become the Sinhalese abl. termination: mâlaṅgin is a compound like Sanskrit mâdṛiça.

^{page 154 note 3} I take ahak as asakta “detached.”

^{page 154 note 4} The etymology of aṇḍinavâ “to weep,” is doubtful.