China 2014

Pam travelled to Yunnan, China in the spring of 2014 to hunt for Primulas. Click image for more...

Primula Rediscovered

Primula bracteata and Primula bullata are found in their type locations after 125 years.

Near Lhasa, Tibet

How do you tell the difference between P. tibetica and P. fasciculata?

Primula ambita in the Wild

The first ever cultivated plant caused a stir at Chelsea earlier this year.

New Primula Book

The latest Primula book is a revision of the 106 species of Primula found in India.

The Mysterious Primula of Omta Tso (Bhutan)

Omta Tso, Bhutan
In November, 2016, I visited the herbarium at the Royal Botanic Garden Edinburgh to study non-type herbarium sheets which are unavailable online. One of the species I wanted to investigate is Primula tsariensis which was described from a collection made in 1936 by Ludlow & Sherriff in the Tsari valley, SE Tibet  28°41'29.98"N 93°23'57.85"E.

It was initially compared with Primula griffithii and P. tanneri but it was distinguished by leaves which are rounded at the apex, bluntly crenulate at the margin and a calyx which is sparsely farinose and glandular. The flowers are described as rich purple.
Primula tsariensis
Going through the herbarium specimens, I came across a folder labelled "P. tsariensis yellow form". What is that?!! All of the collections of this form were Ludlow & Sherriff's from the area near Omta Tso  27°42'54.88"N 90°17'17.96"E in the upper Richen Chu valley, Bhutan. In 1937, Sherriff explored this place and so I traced his 1937 route using his diary and field notes. Sherriff mixed up the names of the two lakes Omta Tso and Thita Tso, so his initial collection #3383 is from the hillsides and shores of Omta Tso (not Thita Tso as stated in his notes). Sherriff gave it the name "Primula strumosa var. nana floribus maximus". Later, W.W. Smith & Fletcher gave it the name "Primula strumosa var. perlata" and in the Flora of Bhutan it is included by that name as a synonym under P. tsariensis. In the genus Primula, there are a number of species which have both purple and yellow forms, so the idea of a yellow P. tsariensis is acceptable.
Primula species of Omta Tso, Bhutan
Continuing to trace Sherriff's 1937 route, I found that he took a two day side trip (July 13-14) from Omta Tso, westwards along the ridge to Chore (I don't have an exact location for Chore, but it would seem to be a yak herder hut at the base of a cliff SW of Omta Tso) and then followed the main ridge south until descending to Maruthang via a steep gully. Sherriff said "I have never seen so many alpines out together as on this march. In places the hillsides and cliffs were just covered with them and the variety was great."
Omta Tso area, Sherriff 1937 - hybrids found along the orange route
Between Omta Tso and Chore, Sherriff came across a little grassy hollow 50 ft. x 50 ft. where there was a swarm of primulas of at least eight shades of colors including yellow, white, purple and blue. These same colors are seen in hybrids where P. calderiana crosses with P. strumosa. Essentially P. strumosa and P. calderiana are yellow and purple forms of the same species, though they have different geographical areas and P. strumosa grows at higher elevations.
Color variations of P. calderiana X P. strumosa
Sherriff knew these were hybrids but couldn't decide which species were the parents. To further complicate matters, Sherriff also found that the yellow, white and blue-purple flowered plants grew in separate masses of one color only. However, the white masses included a few variations of yellow tinged, or blue tinged plants. Sherriff collected what he said was true P. calderiana (L&S 3437) and that it was distinct from the others and not involved in the hybrids. In the field notes, someone (perhaps W.W. Smith) later wrote in the margin "I believe these are normal hybrids of P. calderiana and P. strumosa" and included a reference to hybrid L&S collections from Weitang and Pangotang (Tsampa), Bhutan in 1949, which are unfortunately I haven't seen.

So are the plants at Omta Tso P. tsariensis or are they P. strumosa?

The most obvious differences between the two species are that P. calderiana / P. strumosa smells "disagreeable", and has farina present on the bud scales, and on the upper scape, calyx and pedicels whereas P. tsariensis is not noted for its smell, has efarinose bud scales and only trace amounts of farina on the pedicels and calyx.
Bud scales - (L) no farina, P. tsarieneis, (R) with farina, P. strumosa
Sherriff was limited in his ability to document these plants but certainly more images can shed light on this problem. If you travel to this area and encounter these plants, please take lots of images of the flower color variations, showing the calyx, the basal bud scales, and leaves. And take a sniff!

Pam Eveleigh © 2017

The Black Mountain (Bhutan) - Primulas and Problems

Jigme Singye Wangchuck National Park covers 1,730 sq kms in south-central Bhutan and contained within the park is Durshingla, the Black Mountain. (JSWNP Facebook page)
Durshingla © Darlo Letro
In 1915, Roland Edgar Cooper made the first botanical expedition to Durshingla, which he called Joedawnchi. From Cooper's 1914 and 1915 Bhutan expeditions, 19 species of Primula were described though only six are now considered to be distinct species: P. chasmophila, P. eburnea, P. erythrocarpa, P. strumosa, P. umbratilis and P. xanthopa. He wrote about those trips and specifically about the Primulas he found in Notes of the Royal Botanic Garden Edinburgh, Vol.18, Nov. 1933.

Ludlow and Sherriff (L&S) visited Bhutan several times and Sherriff's 1937 expedition was specifically to Durshingla which he called Dungshinggang. He found no new Primula species there but he was able to recollect known species.

Sherriff's route up Durshingla in 1937
Both expeditions used a route from Chendebi, travelling west over the Lamse La into the Phobjikha valley. From there they ascended the ridge over the Byasu La to Chapepusa and then followed the ridge east until it met the main N-S ridge leading to the peak.

Primulas & Problems

1. The most interesting species found on Durshingla is Primula chasmophila and, so far, it is only found on this mountain. Cooper first discovered it in September and his herbarium specimen was sparse, so it was described from cultivated plants grown from seed. Sherriff collected it on the mountain ridge near the Nabzi La. Though it was scarcely in flower, he noted its deep, rich blue violet color, red eye and its cliff ledge habitat and compared it with Primula umbratilis, a much more widespread species that is also found on the mountain (Cooper 4822). What to photograph if you see this species? Variations in flower color, details of the hairs on the flower scape, leaf study.
Primula chasmophila L&S 3301,
courtesy Royal Botanic Garden Edinburgh
2. The type of Primula hopeana was collected by Cooper on Durshingla, in September, so the plants were in seed. In the description it was noted that sometimes the tube of the flower was "pink", which was seen by Sherriff on the mountain and is seen in populations in Arunachal Pradesh. This species resembles Primula sikkimensis and is sometimes considered a variety of it. Species in the Sikkimensis Section hybridize in the wild and it is very difficult to sort some populations into species. Sherriff noted that he saw true (yellow) sikkimensis here also, so I would expect some hybridizing. What to photograph if you see this species? Leaf study with the corresponding flower color and flower shape. Variations in flower color (don't forget to also show the flower face).
Primula hopeana
3. On the ridge to the NW of the Lamse La and also on the lower ridge of Durshingla Sherriff found a plant he thought was Primula strumosa but was later found to be Primula elongata. Primula elongata, P. barnardoana, P. sikkimensis and P. strumosa are all yellow flowered, so you must look to other characteristics to distinguish between these species. In Bhutan two variations of P. elongata occur, having slightly different leaves. What to photograph if you see this species? The most important image is a leaf study! The flower tube is also important to determine this species (side of the flower).
Primula elongata
4. On the high hills and passes surrounding the Pobsikha valley and lower on the ridge of Durshingla grows Primula whitei. The type location for this species is the nearby Pele La. This is another problem species as it is close to or the same as Primula bhutanica. See this blog post for details. What to photograph if you see this species? Any population of P. whitei should be checked to see if plants have any characteristics of P. bhutanica in particular the calyx (side of the flower). A leaf study, especially of the older leaves, would be useful.
Characteristics of P. bhutanica. Can you find plants like these in a population of P. whitei?
5. Another species in the Petiolares section was found growing in this area (Lamse la, Lao La). It caused Sherriff a lot of confusion as he thought it might be P. boothii, but it was later identified as Primula bracteosa. The Flora of Bhutan has P. boothii as a synonym of P. bracteosa, but there is a distinction between the two in that P. bracteosa develops leafy bracts among the inflorescence as it goes into fruit. In flower, the two species are indistinguishable though it is said that P. boothii does not have farina and has a slightly different calyx. What to photograph if you see this species? It is best if the same plants can be visited in flower and in fruit. The calyx detail and any farina should be photographed. A leaf study of the plants in flower and in fruit, including the later leaves in the inflorescence is necessary.

Primula bracteosa,
courtesy Royal Botanic Garden Edinburgh
6. The first Primula encountered by Sherriff on the way to Durshingla was Primula flagellaris, which he found within a mile of Chendebi, on the right side just before cliffs, on the way to the Lamse La. This has been a problem species because it has also been considered a variety of Primula tenella. The distinctions between P. tenella and P. flagellaris are in the flower size and color, number of flowers per scape (1 or 2), presence or absence of bracts and presence or absence of stolons. I believe the species described in 2004 as Primula rebeccae is a synonym of Primula tenella. What to photograph if you see this species? Both a flower study to show size and color variation and a leaf study (both surfaces) with measurements would be helpful. Images of the stolons (runners) and the new plantlets that form on the ends of the stolons. Flower stalks should be checked for bracts (which look like a tiny leaf) and photographed if found. Sherriff believed there was an altitudinal difference between the two species, with P. flagellaris at a lower altitude than P. tenella. It would be exciting if P. tenella was found at higher elevations on Durshingla.
Primula tenella usually has 1 flower and a bract.
If you are at Chendebi, look for a small species on the near by hills (SW) for Primula erythrocarpa which looks like P. atrodentata or a small P. denticulata. Check for basal scales under the leaves and show the hairs on the leaves.

Other Primula species found on Durshingla: P. atrodentata, P. munroi, P. calderiana, P. megalocarpa, P. waddelli, P. bellidifolia, P. obliqua, and P. glabra.

I am uncertain of the exact location of Nabzi La on Durshingla. If you have the GPS coordinates for it or a GPS track for the route up the ridge, please contact Pam. If you are interested in more details of these species or locations, or if you have images of these species, please contact Pam.

A video of Foresters in Bhutan

Pam Eveleigh © 2017

Spring is here (Primula denticulata)

One of the most photographed Primula in the wild is the showy and locally abundant species, Primula denticulata. It is distributed across the Himalaya from Afghanistan through to Sichuan, but it was first collected in Nepal. It was described in 1805 by James Edward Smith and the original description includes a painting. The type specimen LINN-HS 271.2.2  (Herb Smith) was collected by Dr. Buchanan and resides at the Linnean Society. The type location is "the moist parts of the hills about Chitlong in Nepal, flowering from February to April". Buchanan's journey is detailed in "An Account of the Kingdom of Nepal", written in 1819. Chitlong (27°38'59.27"N 85°10'12.17"E) is very near to Kathmandu, though overlooked by plant enthusiasts as they head for more popular trekking areas.
Primula denticulata from the type location
This species varies across its distribution and authorities have disagreed on the distinctions, creating numerous species which have since been lumped into P. denticulata including P. adenophora, P. aequalis, P. alta, P. cyanocephala, P. harsukhii, P. hoffmeisteri, P. limnoica, P. paucifolia, P. platycrana, P. sinodenticulata, and P. telemachica. A one time even P. capitata and P. erosa were thought to be synonymous. Primula cachemiriana, described from cultivated plants, persists as a name given to plants covered with dense yellow farina. Adding to the confusion are dwarf states which resemble P. atrodentata but can be distinguished by the presence of persistent basal bud scales which were once part of the winter resting bud
P. denticulata basal scales flatten to reveal true leaves and flower buds
Many garden selections have been made. The compact flower heads are composed of numerous erect flowers in shades of purple, pinkish-purple, reddish-purple or white. It is winter hardy in my garden in Calgary, Canada though it does resent water around the resting bud in winter. It is easily grown from seed, which is produced in copious amounts and can be vegetatively propagated by splitting crowns or by root cuttings.

Primula denticulata color forms
Masses of plants in Bhutan
Pam Eveleigh © 2017

J. D. Hooker and Primary Data Sources (Primula pulchra)

P. denticulata, a common species near Lachen, Sikkim
The easiest way to investigate a species is to read secondary sources of information. These are often such works as monographs or regional floras. We are now at least one generation removed from some of the prominent explorers and plant hunters of the Himalaya, such as George Sherriff (1898-1967), Frank Ludlow (1885-1972), Frank Kingdon-Ward (1885-1958) and even further removed from such explorers as Joseph Dalton Hooker (1817-1911). When William Wright Smith and Harold Roy Fletcher wrote a series of articles (see bottom of this book list) in the 1940's on Primula, they effectively wrote a monograph of the Genus. This is a secondary source of information, but the great advantage of their work was that they could personally consult some of the explorers. Now that those explorers are no longer alive, we can't ask their opinion on the distinction between two species, or what the wild habitat was like or even the most obvious question : Where exactly did you discover that species?
Joseph Dalton Hooker (public domain)
Secondary sources can be frustratingly vague on that most basic of questions by copying the location given on the herbarium sheet or in the field notes. Often this location is given as the name of the nearest prominent place to the collection site - not where the specimen was actually collected. Such is the case of Primula pulchra (see my related blog post "Primula pulchra - Hide and Seek with P. gambleana". Hooker gave his unnumbered collection as the type of the species and cited the location as "Lachen, 12,000ft", June 8, 1849. Since Lachen is actually at an elevation of approximately 8,700ft, it is obvious that this is not the actual collection location. So exactly where did Hooker make this collection?

Hooker wrote a two part book called the Himalayan Journals about his travels in this area. Unfortunately this isn't a day-by-day account and so we have to piece together his route by reading several pages starting at pg 47 in Volume 2. We find that he was in the Zemu Valley, and camped at the junction of the Zemu and Thlonok rivers at a place we now call Tallem 27°46'50.48"N 88°29'31.50"E which is at an elevation of approximately 10,700ft - still not high enough to be the collection site. On page 50, Hooker tells us that he repeatedly ascended the north flank of Tukcham mountain (now called Lamo Angdang), but that he also went up the Zemu valley, either of which could get him to the correct elevation of 12,000ft.
Looking from Lachen to the head of the Zemu Valley (L side)
Photo: Abhinaba Basu (Flickr, CC)
Luckily we have access to a primary source of data - Hooker's diary and letters from that time. Thanks to Cam Sharp Jones at the Joseph Hooker Correspondence Project, Kew, I was able to read Hooker's diary entries for June 7th and 8th, but it was still unclear as to exactly where Hooker had ascended on June 8th. Luckily, Hooker wrote a letter to Archibald Campbell on June 9th describing in detail his activities on June 8th. Included is a hand drawn map showing his camp spot which is on the south side of the river, not the north side where modern expeditions camp. He describes his ascent of the mountain (Tukcham) to the south of his camp:

"I went up to nearly 14,000ft by a steep torrent, snowed the whole way up i.e. from 11,000ft up to perpetual snow at 13,500ft, which was there continuous and flanked by lofty black precipes wholly inaccessible. The fatigue of the ascent was very great from the snow, slipperiness, and enormous rocks which are constantly tumbling from above."

There is a gully rising up the mountain from Hooker's camp spot and there is no doubt in my mind that this is where Hooker went. There is an image of this gully
on a website detailing a trip up the Zemu. This a dangerous place to ascend and it is a testament to Hooker that he was able to do it without being injured, especially considering the rock fall. It is likely that Primula pulchra will be found in the nearby Kishong La or up the Zemu valley towards Green Lake, so I have hopes that someone will soon take images of this elusive species.

Pam Eveleigh © 2017

Primula pulchra - Hide and Seek with P. gambleana

In 1882, An article was published in the Journal of the Linnean Society titled "On some Undescribed and Imperfectly known Indian Species of Primula and Androsace" by George Watt. It was actually J.D. Hooker that wrote this article in Watt's name, using Watt's notes, though revising them where needed, and with the intention that this would be used in the Flora of British India that was being written at the time by Hooker.

Primula pulchra or not? (Smith collection)
The first new species of Primula described in the article is P. gambeliana (a spelling mistake of P. gambleana) from a collection (Watt 5483) made at "the Tra Cha Kumpa Kubra Rock (Black Kabur?) above Jongri (Dzongri), Sikkim 27°30'29.99"N 88° 9'9.45"E. It grows in moss and often in crevices of steep wet cliffs.

P. pulchra is the second new species described in the article, from a collection (s.n. 12,000 ft) made by Hooker at Lachen, Sikkim (27°43'2.18"N 88°33'29.02"E, probably in the Zemu valley according to Hooker's Himalayan Journals). Also noted is a syntype collection (Watt 5406) made at the upper Ratong Chu Basin, Jongri (27°32'38.68"N 88° 8'4.47"E) for which Watt 5268 is given as equivalent. This location is very close to the type location for P. gambleana.

Both of these species are figured: P. gambleana - Tab I, and P. pulchra Tab II. A (see image below). Hooker makes the comment under P. gambleana in the Flora of British India: "petiole not sheathed at the base as in P. pulchra, which this a good deal resembles".
Drawings from the Journal of the Linnean Society
If you compare the two drawings side by side, there are a lot of similarities between the two species. Both having a sheathing base, but the scales are much larger in P. pulchra. The most obvious difference is the leaf shape with the leaves of P. gambleana shown as orbicular though Smith & Fletcher later described them as "ovate-cordate or ovate-oblong rather than orbicular" and noted that the figure shows the leaves as more orbicular than is usual. The leaf base is described as usually deeply cordate but occasionally truncate. Smith & Fletcher describes the leaves of P. pulchra as oblong to ovate-oblong with a rounded or slightly cordate base. Both species are shown with an obvious scape with 1-10 flowers. The relative size of the plants on the types specimens are similar.
P. pulchra? capsule (L), P. gambleana capsule (R)
courtesy Royal Botanic Garden Edinburgh
Apparently, the most distinguishing feature is the seed capsule which was not known when the two species were described. P. gambleana has a cylindrical capsule, twice as long as the calyx and dehiscing by longitudinal valves. Seen in only one collection of P. pulchra (Tari, Sikkim, no collector) at Edinburgh is a subglobose capsule equalling the calyx dehiscing by a crumbling wall (Section Petiolares type). This collection is of 6 thin, sparsely flowered plants, and none show the sheathing base that is seen in similar plants in the type collections. Is this really P. pulchra in seed or is it another petiolarid species?
Primula sp Yak Desha = P. pulchra Smith?
In about 1985, George Smith made a collection from the area near Kanchenjunga that was identified as P. pulchra. It was awarded a PC in 1991 and may still be still in cultivation. This clone does not seem to produce the tall scapes of the type collection and it makes me wonder if it really is P. pulchra. In addition to this is a recent image from Yak Desha, Nepal which seem to equate to the Smith collection but again does not seem to be the same as the type collection of P. pulchra!

Perhaps the Smith collection is true P. pulchra, but it is important to get it right because it affects other related species P. chamaethauma and P. chamaedoron which have been lumped into P. pulchra by some authorities.

What is needed is lots of images, including the capsules, of both P. pulchra and P. gambelana from Sikkim in the Ratong basin (near HMI Base Camp) and the cliffs around Black Kabur, near Dzongri and also the Zemu valley. Only with this additional information can this puzzle be solved. If you can help, please contact Pam Eveleigh.

Pam Eveleigh © 2017