in the middle — मध्य is the middle of anything, and the locative of it is how Sanskrit says 'among' and 'between' alike. **The same word stood at 1.18** in the same case. DCS reads मध्ये here as an indeclinable, which it is often treated as; it is the locative of a noun, and the compound before it is a genitive of that noun

and moreover — Mallinātha does not let it be the ordinary conjunction: किं चेति चार्थः, 'the sense of the च is किं च', which is the phrase for adding one more item to a list already running. **The fifth token-stanza and the second strung on a च**, after 2.16. He glosses the particle before he glosses anything else in the verse, because it is what says the list has not finished

board — फलक is a plank or slab, and Mallinātha glosses पीठ, a stool or footboard, so the crystal is the ledge the bird stands on rather than any part of the pole. **The gender change is the third of its kind in four stanzas**, after मार्गा at 2.16 and वृति at 2.18: a bahuvrīhi takes the gender of what it describes, and what it describes here is a feminine yaṣṭi. The masculine फलक is declined as a feminine because the compound describes वासयष्टिः

golden — and Mallinātha derives it the same way he derived the beryl at 2.16: काञ्चनस्य विकारः काञ्चनी, 'a modification of gold', then answers it with सौवर्णी, the same formation on सुवर्ण. **This is the third gold-word of the canto built by that one affix** — हैम and सौवर्ण at 2.16, काञ्चनी here — and the third time he takes the pattern apart rather than name it

roosting-pole — Mallinātha glosses निवासदण्डः, 'a staff for living on', which is exactly what a perch is. **It is the fifth and last of the tokens**, and the smallest: a house, a tank, a hill, two trees, and a bird's pole. And it is the only one that exists for the sake of an animal. He closes the first half अस्तीति शेषः, the fifth stanza running that the sentence has needed a verb from him

at the foot — मूल is a root, and of anything standing upright it is the base. Mallinātha does not gloss the word; what he glosses is the whole phrase, and he glosses it with a noun the verse does not have. **The locative and the participle together mean a built plinth**, and he says so: कृतवेदिकेत्यर्थः, 'that is to say, with a platform made'

set — the past participle of √बन्ध्, to bind, agreeing with वासयष्टिः, with the emeralds in the instrumental as what it is set with. **The same root and the same use as the बद्ध of 2.16's मरकतशिलाबद्धसोपानमार्गा**, where the tank's steps were laid with the same stone. Mallinātha does not gloss the participle either time; he glosses the result — there a stairway, here a plinth

gems — and Mallinātha names them, which the verse does not: मणिभिर्मरकतशिलाभिः, 'with gems, that is, with slabs of emerald'. **He is reading the colour word backwards into the stone.** Nothing in the stanza says emerald; what it says is that the sheen is that of young bamboo, and he takes the green to settle the gem. It is the same मरकतशिला that faced the tank's steps at 2.16. The visarga becomes र् before the अ of अनतिप्रौढ

sheen — प्रकाश is प्र with √काश्, to shine, so a shining-forth, and of a stone it is the light off its face. Mallinātha uses the word twice in his vigraha, once for the bamboo and once for the gems, with इव between them, and then throws both away for a single compound of his own, तरुणवेणुसच्छाय, 'having the same shade as young bamboo'. **छाया is the exact opposite word to प्रकाश** and means the same thing here: a stone's colour is its light and its shade at once

beats — a ताल is a measure of time in music, and Mallinātha says how this one is made: करतलवादनैः, 'by playings of the palm', which is clapping. **The word for the measure and the word for the palm of the hand are the same word**, and the gloss is what makes that visible. The instrumental is what the peacock is made to dance by

lovely — सुभग is 'of good fortune', and of a thing it is what pleases. Mallinātha answers it with रम्य, delightful. **The claps are lovely because of the bangles**, which is to say that the sound of the ornament is part of the music and not an accident of it. The visarga becomes र् before the न् of नर्तितः

made to dance — the causative of √नृत्, so the wife is the agent and the peacock the thing danced. **The third causative participle in five stanzas**, after वर्धितः and नमितः at 2.15, and the third time the household is described by what somebody has done to something that grows or lives there. Mallinātha carries the word over unglossed and only names the agent, मे मम कान्तया

by my wife — the same word in the same case and with the same मे in front of it as at 2.15, कान्तया वर्धितो मे. **She has now done two things in the garden**: raised a tree and taught a bird to dance, and both are said with a causative participle and an instrumental. Mallinātha again writes मम in for the enclitic and leaves the noun alone

my — the unaccented genitive again at the end of its pāda, exactly where 2.15 put it, and Mallinātha again replaces it with the full मम and moves it in front of कान्तया. **Two stanzas four apart end their third pāda on the same word in the same position**, and it is the word that makes the garden hers

which — feminine, so it reaches back to वासयष्टिः across two whole pādas, and Mallinātha writes the noun in with a diminutive of his own: यां यष्टिकाम्. **The accusative is the point.** A perch is where a bird sits, which ought to be a locative; it is an accusative because of the preverb, and he cites the rule for it at the verb

sits on — and this is the verb 2.16 was not. There नाध्यास्यन्ति looked like अधि with √आस् and Mallinātha read it as आ with √ध्यै instead; here he reads अधि with √आस् and proves it with a sūtra: अधिशीङ्स्थासां कर्म, Pāṇini 1.4.46, by which the place of lying, standing or sitting becomes the object when the verb carries अधि — इति कर्मत्वाद्द्वितीया, 'the accusative, because by that rule it is the object'. **The two stanzas are three apart and settle the same-looking verb in opposite directions**, and each settlement is the evidence for the other: the ending here is ātmanepada, which √आस् must be and which the -अन्ति of 2.16 could not be. He glosses it यष्ट्यामास्ते, putting the locative back

at the departing — विगम is वि with √गम्, a going away, and the locative of a time-word is when. Mallinātha replaces the phrase with the plain सायङ्काल, evening. **A peacock roosts at dusk and dances at a cloud**, and the stanza has put both facts in one line without saying that either is a fact

the blue-throat — the word is a bahuvrīhi in shape, 'whose throat is blue', and a bird's name in use, and Mallinātha treats it as the name: he does not divide it, he replaces it, नीलकण्ठो मयूरः, and cites Amara's line of synonyms for the peacock. **It is also one of Śiva's names**, for the throat that held the poison, and the canto has had Śiva in it since 2.14; the verse leaves the coincidence alone and so does he. Same handling as इन्द्रनील at 2.17: no layer here divides it, because neither witness does

friend — सुहृद् is 'good-hearted', a bahuvrīhi frozen into the ordinary word for a friend, and Mallinātha answers it with सखा, **the word this canto has used of the cloud at 2.17 and of the wife at 2.18**. Now it is used of the peacock, and all three friendships are being arranged in the same three stanzas. The peacock is the cloud's friend because it dances when the cloud comes

of you — **plural, where the poem has addressed the cloud in the singular throughout**: त्वाम् at 2.16 and 2.17, तव at 2.18. Mallinātha writes the full form in, वो युष्माकम्, and passes on without remarking on the number. What the plural fits is the fact rather than the addressee: a peacock dances at any cloud, so its friendship is with the kind. The enclitic cannot stand first in its clause, and here it stands last in the stanza