Mānadeva Saṃvat: Old Problem, New Evidence[1]

by Gautama V. Vajracharya

Amendment of Feb. 22, 2022

text and photos © asianart.com and the author except as where otherwise noted

February 14, 2017

(click on the small image for full screen image with captions.)


Fig. 1The Mānadeva and Śaka eras, both of which were referred to in inscriptions made during the Licchavi period (C. 200-879) in Nepal, have been a subject of scholarly debate for over a century.[2] However, within the last two decades, the topic has been examined in greater detail, not only in numerous articles, but also in several voluminous books.[3] Despite heated debate, and a desperate search for evidence to substantiate their own views, a highly significant piece of information unfortunately escaped the attention of international, as well as native, scholars. None of them realized that the Licchavi period inscription carved on the pedestal of the monolithic statue of Śiva and Pārvatī at Sikvabahi (or Siku Bahi) Patan (Fig. 1) is actually the Rosetta Stone; the date of the inscription being recorded there in two different eras, namely, the Mānadeva and the Śaka eras.


Fig. 2In 1969, the Department of Archeology of Nepal discovered the abovementioned inscribed stone sculpture and published a partial reading of the inscription in a local newspaper, Gorakhāpatra (Śrāvaṇa 12, V.S. 2025). According to the department epigraphist, the inscription is dated saṃvat 495. On hearing this exciting news, we, the members of Saṃśodhana-maṇḍala, the well-known institute devoted for investigating Nepalese history, rushed to the Sikvabahi Patan, and after two visits we were able to read the majority of the inscription, which we promptly published in Pūrṇimā, vol. 5, no. 1, 1969. However, the inscribed pedestal of the sculpture was partially buried, and it was difficult to take the rubbing of the last line of the inscription, where the dates are carved, so this published reading was not fully reliable. Moreover, even when my uncle, Dhanavajra Vajracharya, published the inscription a second time, the last line, including the date and its translation (in Nepali), remained both incomplete and inaccurate.[4]


Fig. 3My art historian colleagues, namely, Pratapaditya Pal, at the time, Senior Curator of South and Southeast Asian art at the Los Angeles County Museum of Art and author of multiple monographs on Nepalese Art, and Mary Slusser, the author of Nepal Maṇḍala, took an interest in this epigraph, because the stone sculpture is a rare, early-dated work of the Licchavi period. Both of them discussed the sculpture in detail in their works.[5] In approximately 1971, when I went to see the sculpture with Mary, I found that someone had dug around the statue to a much greater depth than we had managed a few years previously. I was then consequently in a better position to more accurately decipher the last line of the inscription, including the date. (Alas, since the time we read the inscription, the sculpture has been enshrined in a handsome brick structure; but in the process the inscription has been buried in the cement floor up to the upper two lines (fig 3)). Of course, I informed my uncle that the last line of the inscription and translation published in his book required revision. Although he hesitated to accept my reading at first, he did make a partial correction to the last line of the inscription in his hand writing in his copy of the published book, which I have with me. However, for some reason, his corrected version never appeared in the later edition of his book; Dilli Raman Regmi simply copied Dhanavajra’s reading, including his mistakes.[6] Since the correct understanding of the inscription is so crucial, I here not only present my reading, translation, and illustrations of the inscription (Figs. 2 and 4), but also my comments based on an analysis of contemporaneous sources:

• 
  Fig. 4

Text

1. oṃ bahirdeśāddakṣiṇasyāṃ rājaputtravajrarathanāmno …


2. āsīttadgottrajenaikena yāśca vāgvatyā maṇimatyāsca sandhideśe …


3. mātaraḥ sthāpitā āsan mṛṇmayāstāḥ kālakrameṇa cirantanatayātiviśīrṇabhagnapatitapāṇipādā jātā ittthambhūtāśca tā avalokya …


4. … nena ca paścātparamadhārmikena babhruvarmmanāmnāsāmeva mātṛṇāṃpratisaṃskārakriyā cintitāsīdasambhāvyaiva tāṅkriyāṃ sa kāladharmmavaśā …


5. phalopabhuktaye divannītavāṃstadadhunā tadbhrātuṣputrasya nityadharmābhiratacetaskasya paramabhāgavatasya deśavarmmanāmno māttrā pativratayā


6. dharmaniṣṭhayā deśabhaṭṭrikayā tasyaiva kālagatasya babhruvarmmaṇaḥ svarggānantyāya mātāpitrorbhartturātmanaśca puṇyābhivṛdddhaye punaranyathā


7. devyo mātaras sapta śailyaḥ kāritā iti saṃvat varṣaśataṃ 400 90 9 //

Translation

To the south of the place called Bahir, … of a rājaputra [related to a royal family] named Vajraratha … At the confluence of the Vāgvatī and Maṇimatī rivers, one of his ancestors had consecrated the earthen images of mother goddesses. Over time, the arms and legs of the images had broken and fallen apart. Having seen this, pious Babhru Varman decided to repair the images of these goddesses. He died before he could fulfill (his desire), and was lifted up to heaven so that he could enjoy the fruits of his (good deeds). Therefore, Babhru Varman’s brother’s devoted pious wife, Deśa Bhaṭṭārikā, the mother of Deśa Varman, (Babhru Varman’s) nephew, a great follower of the Bhāgavata cult and a serious believer in dharma [religious works], has now commissioned the seven stone sculptures of mother goddesses again, in a different [medium], with a desire to make the deceased Babhru Varman’s heavenly residence eternal, and also to increase the religious merits of her parents, her husband, and herself. [Mānadeva] era 100, [Śaka era] 400 + 90 + 9 [= 499].


Fig. 5Before I explain the reasons for identifying these two eras as Mānadeva and Śaka, respectively, it is vitally important to correctly read the last line of the inscription. A comparison of my reading of the last line and its translations with those of previous scholars, namely the authors of the Department of Archeology, Saṃśodhana-maṇḍala, Dhanavajra Vajracharya and Dilli Raman Regmi, shows that sapta is absent from previous readings, and that the translation of the highly significant compound word varṣaśatam which means “year 100” is also completely missing. More importantly, their reading of the final digit of the number of the second era differs from mine. However, a comparative observation of related numbers found in several Licchavi inscriptions indicates that the number cannot be 5, but instead is 9.[7] These two numbers differ significantly. The number 5 characteristically faces to the viewer’s right, whereas the number 9 faces to the left and appears like stylized earlobe. The last number in the Sikvabahi inscription is obviously facing left. The final line of the inscription runs irregularly, sometimes with a big gap between two words, and other times slanting from left to right. Note that the number 9, as well as the cursive vertical lines, which is actually a punctuation mark called virāma at the end of the inscription, is carved approximately two lines below the level of the preceding number 90. This observation is important because it helps us to realize that the arrowhead like triangular crack, located unexpectedly above the level cannot be part of the inscription but is one of the numerous cracks that we see throughout the inscription. Actually, the crack is isolated and not connected with the numeral. On the other hand, if we consider the triangular crack part of the last number of the saṃvat, the size of the last digit of the saṃvat in combination with the earlobe like numeral becomes irrationally much larger than the preceding numerals and syllables of the inscription. Previous authors did not notice this problem and overconfidently decided that the last digit of the saṃvat is 5. The main problem emanates from the lack of closer examination because of which previous authors could not recognize both the earlobe like numeral and the virāma sign.[8] In fact, the latter is completely missing in their readings. The punctuation mark is not itself that significant, but one should not confuse it with the number of the date either. This critical examination led us to the conclusion that the second era given in the Sikvabahi inscription is actually 400 + 90 + 9 = 499.

Keeping in mind that paleography is an interdisciplinary study, we should now move on to examining how our reading of date is relevant to other contemporary historical events and textual sources. According to an earlier Licchavi period inscription, which is dated Śaka era 482 (561 CE), the deceased Babhru Varman, mentioned in the Sikvabahi inscription as the uncle of Deśa Varman, was an important figure, who was working as a dūtaka for the Licchavi king, Gaṇadeva, and the de facto ruler, Bhauma Gupta.[9] Thus, there is no reason to doubt that the 499 era in the Sikvabahi inscription is a continuation of the era used in the earlier inscriptions of the Licchavi king Mānadeva’s time (c. 464–505 CE) and later. Previous authors have shown that this era is the Śaka era,[10] and I have found no reason to disagree with this view.

More importantly, a careful study of two dates given in the Sikvabahi inscription certainly provides us with evidence enabling the calculation of the difference between the Śaka era and the newer era. The inscription tells us that the latter reached 100 (varṣaśatam) when the Śaka era reached 499; hence, their difference is 399. In other words, the newer era began in year 399 of the Śaka era. King Mānadeva reigned from around Śaka era 386–427 (c. 464–505 CE). Thus, it becomes abundantly clear that the era 100 used in this Sikvabahi inscription began when King Mānadeva was in power. Although we do not know exactly why this new era was introduced in year 399 of the Śaka era, this was undoubtedly the Mānadeva era. As we will see shortly, the Mānadeva era was actually initiated in year 398 of the Śaka era (476 CE); but there is a reason why in the Sikvabahi inscription the Śaka era is 499.

The question now is how our study of the double dates of the Sikvabahi inscription relates to the celebrated reference to Mānadeva saṃvat in the colophon of the Sumatitantra manuscript. In his analytical study of the colophon, Guru Nayarāja Panta showed that the difference between the Śaka era and Mānadeva saṃvat was 498 years. He was of the opinion that the newer era, namely the Mānadeva era, began in CE 576 (Vikrama saṃvat 633). This view is convincing, because his analysis of the date is also in accordance with Tibetan documents, which refer to a Nepalese era that began in CE 576.[11]

At first glance, these Nepalese and Tibetan sources may appear to conflict with my reading and calculation of the two eras, which indicates that the Mānadeva era began in 476 CE, rather than in 576. However, an observation of an ancient system of reckoning annual years suggests that these other sources actually substantiate my finding.

Ancient Nepal was familiar with the Kuṣāṇa period (1st century bce – 3rd century ce), in which art and culture prevailed in Mathura. We know this from Jaya Varman’s statue and the inscription on its pedestal.[12] This statue bears multiple stylistic similarities to the Kuṣāṇa/Mathura images. The erection of a statue of a king was a new concept in itself, and originated from Kuṣāṇa/Mathura shrines that contained a monolithic portrait statue of the Kuṣāṇa king. However, more closely related to our discussion is the style of date enumeration, which included the name of a season and one of the 4 months of that season in a number, with no referral to the names of the months. This system sharply contrasts with the later system, which does not mention the name of the season, but provides the name of the month, followed by the days of the dark half or the bright half. The inscriptions of Kuṣāṇa Mathura are marked by an archaic way of reckoning the era, by which, when the era reaches 100, the next year is not 101, but is 1. Although this system of omitting 100 was known to Alberuni as lokakāla (a common local practice of reckoning time),[13] it was almost abandoned in India during the Gupta period. However, in Nepal, perhaps because the culture of the Kathmandu valley characteristically takes a great interest in keeping earlier traditions alive (which can be proven in numerous different ways),[14] the system was not completely forgotten, even in year 499 of the Śaka era, when the inscribed Śiva Pārvatī image of Sikvabahi was established. The following year of the Mānadeva era when it reached 101, it was treated as a new era, following the tradition of omitting 100. This is the reason why Tibetans designated this omitted 100 Mānadeva era as a new era established by Aṃśu Varman, who rose to the power around that time and gave preference to this new era in his inscriptions. The author of the well-known additional statement given at the Sumatitantra colophon continued to call this era the Mānadeva saṃvat. Consequently, the difference between this Mānadeva era and the Śaka era is no longer 398 or 399, but is 498 or 499 years depending upon the month of the year. Although there is no contemporary evidence to prove that a king named Mānadeva ruled around this time, modern historians, including Luciano Petech, Nayarāja Guru and my uncle Dhanavajra Vajracharya, perhaps erroneously believed that the Mānadeva saṃvat mentioned in the colophon must have been established by Mānadeva II. In fact, the Mānadeva saṃvat described in the Sumatitantra manuscript and the Mānadeva saṃvat of the epoch era that began during the time of King Mānadeva (r. c. 464– 505 CE) are one and the same. It became a new Mānadeva saṃvat only because the 100 years were omitted as it turned into 101. After this time, the custom of omitting 100 was apparently abandoned and eventually forgotten. This has created confusion, even among modern historians who did not pay much attention to the Kuṣāṇa/Mathura inscriptions, and the incorporation of Mathura culture in Kathmandu valley culture during the Licchavi period, until the discovery of Jaya Varman’s statue.[15]

Finally, we must explain that if our reading and translation of the Sikvabahi inscription is correct, and is in accordance with the Sumatitantra colophon, then the difference between the new Mānadeva saṃvat and the Śaka era is presumed to be 498, rather than 499, years. This can be satisfactorily explained. Mānadeva saṃvat was Kārtikādi; it started in the month of Kārtika. This view is based on the Licchavi inscription found near Anantaliṅgeśvara temple, which refers to the annual rite performed on the 11th day of the bright half of the Kārtika month as the very first rite of the year. This does not mean that the Śaka saṃvat was also Kārtikādi. As in India, it must have begun in the month of Caitra. In his Changu Narayana inscription, Aṃśu Varman used an interesting term, svasaṃsthayā (in accordance with the local tradition), apparently, to differentiate the Mānadeva and Śaka eras.[16] In fact, Mānadeva saṃvat was a ritual era. Local festivals and rituals were celebrated both in Nepal and Mathura during the Kuṣāṇa period, in accordance with the annual custom that starts from autumn, either in the month of Āśvina or Kārtika, although the official dates of the inscriptions follow the Caitrādi Śaka saṃvat. This we know from the Mathura inscription made during the reign of the Kusana monarch Kaniṣka, which refers to Mahānavamī of Caitra Dasaim as [u]ttarāyāṃ Na[va]mikāyām “on the day of second Navamī”.[17] In fact we know for sure that the system of running new year (saṃvatsara = period of gestation) in the month of Āśvina or Kartika is related to the life style of the cow breeders of Vedic and pre-Vedic South Asia including Nepal.[18]

According to Śaṅkaramana Rājavaṃśī, the era used by Mānadeva and other early Licchavi kings is actually the Licchavi era; it starts from the month of Kārtika. He argued that the Āṣāḍha month comes before the Kārtika month in a Caitrādi era; however, King Mānadeva was alive when Nirapekṣa’s Changu inscription dated Kārtika at 427; the king must have died before the month of Āṣāḍha 427, because in her Suryaghat inscription of that date, Princess Vijayavatī, described him using the past tense (bavhūva). At first glance, this argument may appear to be convincing, but we must examine it carefully.

Both of these inscriptions, like other inscriptions of the Licchavi period, were written in Sanskrit. We know for sure that Sanskrit was no longer a spoken language around this time; since Sanskrit was the second language of the authors of the Licchavi inscriptions, we find the reflection of their mother tongue in their writing, which is similar to the Newari language, in which the past tense is frequently used to describe the present. Thus, the past tense (aramayat) was used to describe Mānadeva in not only the Suryaghat inscription, but also in the Lajimapat inscription made by Mānadeva’s queen, Kṣemasundarī, which is dated Jyeṣṭha 390. Perhaps, the most noteworthy is the fact that the author of Vijayavatī's Sūryaghat inscription uses past tense (abhūt) even for Vijayavatī. As a donor of the Suryaghat Śivaliṅga, she was certainly alive then. Nayarāja Guru rejected Rājavaṃśī’s view that the era used by Mānadeva and his successors was the Licchavi era. However, due to the lack of any solid evidence, he reluctantly accepted that, just like the Mānadeva era, the Śaka era was also Kārtikādi. Unfortunately, when Nayarāja Guru was alive, the Sikvabahi inscription was not correctly deciphered, and consequently its significance was not properly understood either.

The Sikvabahi inscription actually provides us with a clue to understanding that, unlike Manadeva samvat, the Śaka era used in the Licchavi inscriptions was not Kārtikādi. Because the double dates given in the inscription record the time difference between the Mānadeva and Śaka eras as being 399 rather than 398, it becomes possible to deduce that when the Mānadeva era was initiated in the month of Kārtika the difference between the Mānadeva and Śaka eras was 398. Keeping in mind that Manadeva era went through the system of omitting hundred when the era completed a century, it may not be too difficult to understand that this difference is in harmony with the Sumatitantra colophon. But in the month of Caitra, Śaka era turned into 399. As a result, the difference between them is no longer 398, but 399.  Similarly, in the month of Kārtika of Śaka era 499, when Mānadeva era became 101, the difference between the two eras was 398. Prior to Kārtika of the same year namely the Śaka era 499, when the Mānadeva era was 100, the difference was 399. Thus, we know for sure that the Sikvabahi inscription must have been commissioned during the period of 7 months, from the bright half of Caitra to the bright half of Kārtika, when the difference between these two eras was 399. With this observation, we know for certain that the Śaka era in the Licchavi inscriptions was Caitrādi.

Conclusion

On the basis of the Sikvabahi inscription, we can safely conclude that the Mānadeva saṃvat mentioned in the Sumatitantra colophon is actually a continuation of the era established by King Mānadeva, rather than King Mānadeva II. The difference between the Mānadeva and Śaka eras was originally 398 or 399 years, depending on the month of the year. It was only because Mānadeva saṃvat was considered a new epoch era after it completed 100 that the variation increased by a century, and the difference between them became 498 or 499. The Mānadeva era was known to Tibetans as the Aṃśu Varman era, because it gained popularity during Aṃśu Varman’s time. It started from the month of Kārtika, whereas the Śaka era, just like in India, started from the month of Caitra. This is why the difference between these two eras was 399 when the Sikvabahi inscription was carved.

Gautamavajra Vajrācārya is a Sanskritist and scholar specializing in the art and culture of South Asia. As a young Newar scholar in Kathmandu, he assisted Mary Slusser when she was writing Nepal Mandala. He received a prestigious Rockefeller grant to work as a trainee at the Los Angeles County Museum of Art with Pratapaditya Pal, PhD, Senior Curator of Indian and Southeast Asian Art, on the first exhibitions of Nepalese art. Vajrācārya later earned a Master of Art degree in Art History from the Claremont Graduate School, California, and a PhD in South Asian language and culture from the University of Wisconsin, Madison. He has taught Indian art and culture at the University of Wisconsin for more than two decades. Among his most recent publications is the book Frog Hymns and Rain Babies: Monsoon Culture and the Art of Ancient South Asia. He is guest curator for the exhibition Nepalese Seasons: Rain and Ritual (May 6, 2016 – March 27, 2017) at the Rubin Museum and author of the related catalogue of the same title.

Footnotes

1. A slightly different version of this article is also published in SINHAS, vol. 21, no. 2, December 2016.

2. Bhagavanlal Indraji, Twenty-three Inscriptions from Nepal, together with Some Considerations on the Chronology of Nepal. (Translated from Gujarati by G. Buhler, reprint from Indian Antiquity Vol. 9, August 1880), Bombay: Education Society Press pp. 163-194.

3. In addition to Indraji’s work, two other important articles have shown the gradual development of the multiple theories on Licchavi saṃvats: K. P. Jayaswal, “Chronology and History of Nepal 600 B.C. to 880,” Journal of the Bihar and Orissa Research Society, vol. XXII, pp. 157-264. (Published in the book-form by Bharati Prakashan Varanasi, 1976), and Luciano Petech, “The Chronology of the Early Inscriptions of Nepal,” East and West, vol. 12, no. 4, 1961. The most exhaustive and in-depth analysis of this subject was provided in the following book in Nepali: Nayarāja panta, Devīprasāda Bhaṇḍārī, Keśavacandra Nyaupane, Licchavisaṃvatko Nirṇaya, Kathmandu: The Royal Nepal Academy, 1985. After its publication, two Nepalese authors responded to the study on two different occasions: Śyāma Sundara Rājavaṃśī, in Licchavikālīna Saṃvatko Nirṇaya, Kathmandu: Girīndramāna Rājavaṃśī, 1995; and Kamal P. Malla, “Mānadeva Saṃvat: An Investigation into an Historical Fraud,” Contributions to Nepalese Studies, vol. 32, number 1, January 2005, pp. 1- 49. Rājavaṃśī’s arguments were based on his new discovery of the Gokarṇa Baluva Tar inscription, dated saṃvat 536 dvitīyapauṣaśuklapañcamyām. This discovery is important because it reveals that, although King Aṃśu Varman preferred Mānadeva saṃvat, for which there may be a hidden explanation, he did not totally abandon the use of the Śaka saṃvat. However, this discovery does not support the author’s central thesis that the so-called Mānadeva saṃvat is an indigenous Licchavi era used in later period with omitted hundred. The view of K. P. Malla, expressed in his article “Manadeva Samvat…,” is primarily inspired by Rajavamsi’s Nepali publication.

4. Dhanavajra Vajracharya, Licchavikālakā Abhilekha, Kirtipur: Institute of Nepal and Asian Studies, 2002, pp. 211-213.

5. Pratapaditya Pal, The Arts of Nepal. Leiden: E. J. Brill, 1974, pp. 25, 77, 87-89. Mary Slusser, Nepal Mandala. 1982, 2 vols. Princeton: Princeton University Press, pp. 233-234. As we know from Sanskrit texts on Hindu iconography, it was customary to establish an image of Śiva together with the images of mother goddesses in a shrine (Jitendra Natha Banerjea, The Development of Hindu Iconography, New Delhi: Munshiram Manohrlal Publication, 1974, p. 505). An example of such a shrine sheltering third century CE stone sculptures of mother goddesses, together with the image of Śiva and Pārvatī, is still intact in front of the Bhagh Bhairava temple in Kirtipur. Thus, the reference to seven mother goddesses in the pedestal inscription of the Śiva and Pārvatī’s image at Sikvabhi clearly indicates that this image was originally accompanied by a set of seven images of mother goddesses. Although they are now missing, it is very likely that they remain buried in the ground around the shrine.

6. D. R. Regmi, Inscriptions of Ancient Nepal, 3 vols. New Delhi: Abhinav Publications, 1983, vol. 1, pp. 46-47; vol. 2, pp. 27-28. Although Regmi’s reading was incorrect, the photograph of the rubbing of the inscription that he published in his work (vol. 3, plate no. XLVIII) is helpful. Both his photograph and Mary Slusser’s photograph of the inscription helped me to correctly read its last line. Unfortunately, at the present time, the most part of the inscribed pedestal is again covered with cemented pavement (see fig. 3).

7. Readers may also find it interesting to examine Rājavaṃśī (1974: plates 94–101) and Panta et al. (1985: 114, 247, 271–273, 276–278, 303, 507). Both works contain eye-copies of the numerals of the saṃvat found in the Licchavi inscriptions. Naturally, eye-copies are conjectural and frequently differ from what is real. For instance, compare Rājavaṃśī’s eyecopy of Sikvabahi saṃvat with the photographs of the saṃvat that we have provided here (fig. 4).

8. Such a punctuation mark also appears in other Licchivi inscriptions. For example, a similar mark can be observed in the Jyabahal inscription dated saṃvat 535. (Vajracharya 242-343; Raniero Gnoli, Nepalese Inscriptions in Gupta Characters, part 1,Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1956, inscription 40, plate XLI.

9. Vajracharya, Licchavikālakā Abhilekha, pp. 192-197.

10. Nayarāja Panta, Licchavisaṃvatko Nirṇaya, pp. 183- 191.

11. Petech, “The Chronology of the Early Inscriptions of Nepal,” p. 228.

12. Castro, Angel Andrea Di and Riccardo Garbini, "An Inscribed statue of year 207 from Maligaon, Kathmandu." East and West (Rome: IsIAO), 46: 3-4, December 1996, pp. 299-317; Kashinath Tamot and Ian Alsop, “A Kushan-period Sculpture from the reign of Jaya Varman, a. d. 185,” Asianart, Santa Fe, July 10, 1996, http://asianart.com/articles/jaya/.

13. Edward C. Sachau, Alberuni’s India, 2 vols. Delhi: Low Price Publications, 1989, vol. 2, p. 8.

14. Gautama V. Vajracharya, Frog Hymns and Rain Babies: Monsoon Culture and the Art of Ancient South Asia, Mumbai: The Marg Foundation, 2013, pp. 142-201; Nepalese Seasons: Rain and Ritual, New York: Rubin Museum of Art, 2016, pp. 8-22.

15. Riccardo Garbini, “On Dating the Jaya Varman Inscription: Two Hundred and Seven of the Śaka Era?” East and West, Vol. 52, No. 1/4 December 2002, pp. 421-426.

16. Vajracharya, Licchavikālakā Abhilekha, pp. 317-319. Compare svasaṃsthayā with svasthānī, which also means “associated with locality” as in svasthānī-vrata.

17. Heinrich Lüders, Mathura Inscriptions. Göttingen: Vandenhoeck and Ruprecht, 1961, pp. 208-209. Lüders mistakenly believed that uttarā navamī was the name of a locality.

18. Vajracharya, Frog Hymns and Rain Babies, pp. 29-32, Nepalese Seasons, p. 21.