The report is reproduced by permission from the copyright owner, Brathay Exploration Group.
List of members Itinerary Aknowledgements General Information Mykines Summary of work Ringing Report Storm Petrels & Leach's Petrels
List of Members
Peter Mawby St.John's School Leatherhead Roger Lovegrove Newtown High School, Mont. Geoff Morgan Ysgol Preseli Pembs. Ian Walker Maidenhead Peter Tate Shrewsbury School Paul Watson U.C.N.W. Bangor Stuart Barkway Stanton and Staveley Simon Boyd Clifton College Ken Chorlton King's School, Macclesfield Johnathon Hardey Sedbergh School Keith Hasler Drake and Scull Engineering John Major Sheffield Univ. Patrick Matthews Shrewsbburry School Bill Nash Massey Ferguson Michael Timmis Shrewsbury School
July 20th. Dep. 18.30hrs from Glasgow Airport by Icelandair flight to Vagar (Faeroe Is.) Arr. Vagar 20.30hrs and established temporary base in old schoolroom at Sørvag. 21st Day spent checking equipment etc. and then exploring local areas for birds/plants. 22nd Left by boat for Mykines. Arr. 13.00hrs and establish camp. 31st. Peter and Gillian Mawby arrived. Aug. 5th. Ken Chorlton broke his leg. He and Peter Mawby taken to Torshavn by Royal Danish Navy. 7th. Left Mykines. Re-established base in schoolroom at Sørvag. 8/9th Time left free for members to make the journey to Torshavn - shopping etc. 10th Ken Chorlton transferred from hospital, to Sørvag. Party left by air at 14.45. Arr. Glasgow 17.15hrs.
Our sincere thanks are due to the following people who helped in diverse ways to make the expedition successful.
The council of the Exploration Group and A. B. Ware, Director of Brathay, for their initial help and encouragement.
Peter Weihe of Sørvag, who made our way smooth on Vagar and helped in a multitude of ways.
Leon Heinesen, lighthouse keeper on Mykinesholm, and his wife Erika, who gave us hospitality and endless help with various strange requests.
Olle Niclassen, boatman, Mykines, who gave us a memorable trip by sea round the island, and gave us unparalled views of the island bird cliffs.
Dr. M. O. Preuss, of the_Zoological Museum in Copenhagen for his assistance with the supply of rings etc.
David Sanders, lately warden of Skomer, for help with information on the house mice and for the loan of equipment.
Ivor Rees Esq. Marine Science Laboratory, Menai Bridge for his analysis of food samples for us, from storm and Leach's Petrels
Aneurin Jones, Preseli School, for skilled assistance with illustrations.
Misses Williams and Phillips, Preseli School, for invaluable assistance with a lot of typing and duplicating.
By any Faeroese standards the summer of 1968 was memorable; within the recollection of many of the islanders, the finest,driest summer on record. It was our good fortune to enjoy three weeks of near perfect weather. It rained the whole day on 21st. July - the day we were on Vagar prior to going to Mykines - and during the night of 7th./8th. Aug. when most of us were in Torshavn. During the time we were on Mykines we had virtually no measurable rain (see met. record). As the met. record shows, many of the days were fine warm and brilliantly sunny and the nights scarcely less warm.
Mist and fog are very common features of the islands in Summer. Sometimes it reduced visibility at sea level for several days on end (e.g.26th.-30 July) and mist habitually covered the top part of Mykines and the other visible islands. It was on exceptional day when we could work on the moors above about 300 m. and see further than 100-200 yards.
Winds were light throughout our stay, but variable in direction a1though commonest from the southerly quarter, Which is apparently usual in summer. It was due to a succession or virtually windless nights that our mistnetting of petrels went on night after night,only two nights being missed.
A general map of the Faeroes (scale 1 :200,000) is available and was very useful. A larger scale map (1: 20,000) is also available and was indispensable to us. This latter series covers the whole of the Faeroes although four or five maps are necessary to cover some of the larger islands. All are available from Edward Stanford, Long Acre, London E.C.2 but one has to state the purpose for which the large scale map is going to be used. Mykines is at present being resurveyed.
Standard Brathay supplies were taken and were thoroughly adequate. The only food bought locally was bread and butter and occasionally eggs and milk. Fish was available from the local boats, usually for the asking. The local shop on Mykines sells basic foods - flour, sugar, etc. but not bread which we had sent out on the post boat from Sørvag twice a week. Shops in Sørvag provide almost anything likely to be needed.
There is a regular service between all the islands and linking most of them directly with Torshavn. Land transport is by hire car or bus, both of which are easily available - cars however are expensive. The post boat links Mykines with Sørvag on Vagar on Wednesdays and Saturdays, weather permitting and takes passengers. This is the only regular passenger link with the rest of the archipelago.
Camp Sites - Mykines
There is no good camp site on Mykines, but several alternatives all of which have their various disadvantages. We camped by the stream near where it drops into the sea by the harbour. It was fairly sheltered but water had to be carried from the piped supply in the village. The stream below the village is an open sewer, unfit even for washing hands etc., and allablutions had to be carried out upstream of the village (1/2 ml).As mentioned in the itinerary, we stayed in the old school room at Sørvag during our few days at either end of the trip, and this is an ideal base on Vagar.
Health generally was very good. The only troubles were one or two short-lived cases of sickness and diarrhoea, one of heat rash and one of toothache.
On 5th. Aug. one of the party - Ken Chorlton - broke his leg when he dislodged a boulder while scrambling at the base of the cliffs near the camp site. He was taken on improvised stretcher to the camp and the emergency services were alerted to get him to hospital at Torshavn. A Royal Danish Navy vessel came out from Torshavn (4 hrs) and after a difficult portage and transfer to dingy and then to naval vessel at sea, he was taken into Torshavn and reached hospital 11 hrs. after the accident occured. He returned with us on the flight from Sørvag five days later.
Daily Routine and Camp Organisation
The nature of our nocturnal work with the petrels resulted in a somewhat abnormal daily routine. The weather was so good that mist netting went on night after night; on only two nights was the wind sufficiently troublesome to prevent us putting up the nets. Some of the party had only two or three nights off out of the seventeen we were on the island. Thus the daily routine evolved along the following lines:
Usually about two-thirds of the members would be engaged in a night's ringing on the Holm and so the remainder, that is the fatigue party for the day and some "mouse project" men, plus one or two others concerned with one of the field studies, would carry out 8 more normally timed routine for that day, probably joining in again the following night to take their turn once more with the petrels. So, usually a good percentage of the day was spent in the field (13.00-19.00,22.00~03.30 = 11 hrs.) In the end this, plus the upset to the normal, conventional routine resulted in the need for two days being used as complete rest days for some of those members who had spent many nights on the Holm.
c.11.30 Rise noon "Breakfast" l3.00 Start the days work 19.00 Dinner 20.30 Net-erecting party depart for the Holm 22.00 Ringing party to the Holm 03.00 Ringing party return from Holm 03.30 Bed
Mykines is the most Westerly of the Faeroe Islands (7°35'w: 62°6'N) lying some three miles west of the larger island of Vagar. It is six miles long on its E-W axis and about two miles wide at its widest point. Geologically it is the oldest of the islands in the archipelago made up of a hard, fine-grained anamesite basalt. It is completely cliff-bound and exhibits classically the chief coastal feature of the islands; the succession of exposed lava flows forming rocky walls, alternating with grassy slopes of talus below each. At one point on the South cliffs thirteen such "steps" can be seen. Similar inland cliffs are also a feature of the higher, Eastern half of the island.
Mykinesholm, a small islet half a mile by about a quarter mile at the western end of Mykines, is linked to the main island by a (foot) suspension bridge. It has a lighthouse and the accompanying keepers accommodation, but the light is shortly to become automatic and the accommodation demolished.
The community on Mykines comprises about 60 people, but it is noticeable that there is a population gap of men and women in their twenties and thirties, in common with so many of the remote Western communities off our own islands.
Mykines, and more particularly the Holm, has long been renowned as a sea-bird station of immense proportions, and well it deserves its fame, for the number of Puffin, Kittiwakes, Fulmars, Terns and Guillimots amongst others which breed here, almost defies belief.
These hordes of birds are harvested annually by the islanders as in the other islands, and are salted or frozen to provide winter meat. The islands are in the path of one of the branches of the Gulf Stream and therefore the climate is mild, but often wet, stormy and foggy.
Summary of work
The main aim of the expedition was to work on the colonies of Leach's and storm petrels and manx shearwaters on the Holm. It was also hoped that time would permit other wider biological field studies to be attempted in varying degrees as opportunity arose. The following studies were carried out and formed the main work of the expedition: 8 report of each follows this summary.
1. Ringing: mainly petrels,but also several other species of sea birds. 4172 birds were ringed mainly storm petrels(1881),Leach's petrels(1019) and Arctic Terns (1100).
2. A large amount of data was collected from the two species of small petrels caught. This has been analysed in outline and it is hoped to incorporate it in a paper on the small petrels.
3. A daily log of bird observations was kept and is produced here as the Systematic Bird Report, and includes where possible estimates of numbers etc.. The whole island was covered and we believe that all breeding species were located.
4. A quadrant survey was done of part of the enormous Lambi puffinry to give a numerical sample. An attempt has been been made to relate this sample to a larger part of the puffinry.
5. Mykines contains a distinct sub-species of Housemouse. Daily trapping took place and the results are discussed in comparison with the only(?) other work done on this mammal - by Degerbøl.
6. A list of plants for the island1 was compiled and most parts were visited for this reason alone. Whereas it is not claimed that this is by any means complete it is felt that it is reasonably comprehensive. Apart from the studies enumerated above, other lines of enquiry were followed, pursuing individuals interests etc.. These are summarized in the report and represent limited preliminary surveys mainly in the hope that they may be of use to those who take part in any future expeditions.
7. Meteorological report.
8. Geology - limited preliminary surveys. A collection of samples from Mykines has been deposited at Brathay.
9. Soil profies were taken in a few representative areas and a typical one is included. Also a transection the peat depth on the turbaries was taken and compared to one also taken on the adjoining moor.
10. An entomological collection was made by one of the boys, working on his own and it is hoped that this will eventually be documented and can then be referred to at Brathay.
The following ringing list is virtually self-explanatory.
Wonderful weather conditions night after night account for the high catch of small petrels. The nights were seldom completely dark and the hours of the catching extended only from approx. 10.30 p.m. - 2.15 a.m. It was interesting to see these birds, particularly the Storm Petrels, flying in crepuscular light, when they would certainly not have been flying at home. Only on two nights was the wind (force 3-4 or thereabouts) sufficient to prevent us netting.
We ran out of Leach's Petrel and Arctic Tern rings early on and had to telegraph for more from Copenhagen, which arrived, luckily, very quickly, but not before we had had two nights of throwing away Leach's.
The total of other sea-birds ringed could easily have been higher (particularly Fulmars) but we preferred to devote more of the daylight hours to the other branches of the work. The totals however also contain 22 bonxies ringed by Peter and Gillian Mawby on Svinoy and at Saksun.
Some ringing recoveries
431635 Lesser Black-backed Gull ringed by B.E.G. Sandoy 23/7/66 recovered Ponteverda, Spain ca. 10/2/68 431605 Lesser Black-backed Gull ringed by B.E.G. Streymoy 2/7/66 recovered Algarve, Portugal, 10/11/66 431661 Lesser Black-backed Gull ringed by B.E.G. Streymoy 29/7/66 recovered 37'00 N, 06'20 w Spain 1/11/66 502606 Oystercatcher ringed by B.E.G. Sandoy 9/7/66 recovered Castle Douglas, Scotland, 21/4/67 666738 Storm Petrel ringed by B.E.G. Foula 9/8/67 caught and released (B.E.G.) Mykines 29/7/68 666297 Storm Petrel ringed by B.E.G. Foula 27/8/67 caught and released (B. E. G.) Mykines 28/7/68 9196434 Storm Petrel ringed by B.E.G. Mykinesholm 00.30hr. 27/7/68 controlled Fair Isle Bird Obs. 00.30hr. 31/7/68
Table of ringed birds on Mykines
Date July 23 24 25 26 27 28 29 30 31 Aug 1 2 3 4 5 6 7 Totals Storm Petrel 62 169 149 146 109 177 13 97 135 236 228 1 330 29 1881 Leach's Petrel 109 136 128 136 83 82 11 131 61 57 55 30 1019 Fulmar 1 1 1 4 1 19 2 5 29 63 L.B.B.Gull 5 1 6 Black Guillimot 3 1 4 Puffin 1 2 1 1 2 1 1 9 Manx Sheerwater 1 1 3 2 1 1 1 6 7 1 4 3 31 Arctic Tern 400 522 57 58 63 1100 Kittiwake 7 1 5 13 Great Skua 2 22 24 Arctic Skua 9 1 9 19 Eider Duck 7 1 Oystercatcher 1 1 2 Totals 171 315 279 294 601 270 26 232 210 264 9 814 58 476 130 23 4172
Storm Petrels & Leach's Petrels (R. R. Lovegrove)
Both petrels breed in great numbers on Mykinesholm and reputedly (Williamson 1947) on the western part of Mykines itself; we certainly found the storm petrel here in fair numbers, but saw no Leach's petrels and only heard one or two flying. In this respect however, Leach's petrels - and Manx Shearwaters - were often to be heard calling over the camp site by the village at night.
On the Holm there seems to be a great amount of overlap between the two species; often they can be found in alternate burrows. At the same time there are places where a clear line can be drawn. The best example of this is on the hog's-back ridge which runs from Klettur westwards along the top of the sheer north cliff. On the seaward side there is a short and precipitous slope of loose earth and stone before the cliff drops sheer into the sea. On the south side of the ridge grassland runs right up to the top. Here one can walk in the perpetual half light of the summer lights and see the swarm of storm petrels silhouetted against the sea and the night sky on the north of the ridge as they flit in and out and over the burrows and crevices, while in the dark on the South side Leach's petrels call invisibly over and from the grassy burrows on their side of the ridge. Most of our catching at night was done along the length of the south side of the Holm where the colonies are thoroughly mixed. Both birds occur the whole way from the bridge to the point of Bølid, opposite the two gannet stacks.
A mass of data was recorded from as many of the birds as we could manage. Sometimes birds were released without the data being taken in order to enable the recorders to catch up on the numbers being brought to them. The following details were recorded and are explained in the respective sections dealing with the analysis of this data;
Ring No. Age/sex Weight(Gms) Wing length Pale Race? Leg injury? Regurgitation? Brood Patch
One of the aggravating things about these highly intriguing birds is that it is, on our present knowledge of the species, apparently impossible to sex or age them even in the hand. Thus although about 3000 petrels were handled barely any were sexed(one laid an egg while being ringed; one or two others had just laid eggs) and none was "aged". This inability becomes an added impediment to the work as it holds the answer to one or two interesting problems concerning the birds. Far example, British storm petrels are known to occur in tropical and even S.Atlantic waters in winter and yet some also stay in the N.Atlantic, as witness the occasional "wrecks" on the western seaboard of Europe; what populations or what age groups go where? Again, the bigger colonies of storm petrels are characterized by sudden increase in numbers in June and July. What are these "visitors"? Experienced breeders not breeding each year? Immature birds? Or what?
We hoped that by the study of many hundreds of birds in the careful recording of the details mentioned above we might perhaps formulate some theories on ageing and sexing, especially as the status of some of them was bound to be known, but nothing evolved and neither has anything come in this direction from the analysis of these details.
State of Breeding
Leach's petrels were clearly at a more advanced stage of breeding than the stormies. Eight Leach's burrows that were opened contained five chicks (1/4 - 1/2 grown); one addled egg still being incubated; two deserted eggs (this years?). Furthermore two other factors point to the majority of breeding pairs having reached the chick stage. Few birds had vascularised broad patches (a usual sign of incubation) but many of the birds caught regurgitated food intended for young.
Three storm petrel burrows were opened and all three contained adults incubating very fresh eggs. One stormy, as mentioned above, actually laid an egg whilst being handled for ringing (29th. July) and several others showed signs of having very recently laid.
Storm petrels, like Leach's, regurgitate food readily when feeding young and it should be noted that only six birds out of 1881 handled were recorded as regurgitating food. Thus the conclusion, that we reached is, that the stormies were still lying or in the early stages of incubation, at least several weeks behind the Leach's. In fact some of the stormies at least, were a clear 7 weeks behind some of the Leach's, which were feeding young about three weeks old (by 29th July). Thus these Leach's must have laid at the end of May or the first few days of June which, in fact coincides with the period given in the "Handbook" et al., for British breeders.
Arrival and Departures Times
The storm petrels were always flying over the colony at night before the first Leach's. On no occasion did we positively see a Leach's petrel first. Occasionally a Leach's would be heard on the wing before being seen, but the storm petrels were always silent.
The following table gives the arrival times of both species on those nights when we recorded them accurately. We also took a set of departure times but as the specific identification was not always noted with certainty, they are not included.
Table of arrival Times-Small Petrels
Stormy Leach's Weather 23 July 23.15 23.40 Calm, very light 24 23.05 23.26 Calm, clear 25 22.49 23.05 Calm, clear 26 22.50 23.00 Misty,overcast 27 - - - 28 22.25 22.38 Thick mist 29 - - - 30 23.20 - Calm, light 31 22.53 23.11 Light wind,misty 1 Aug. 22.57 23.15 Light wind, clear 5 22.30 22.50 Near calm, overcast 6 22.46 - Calm, overcast
This table shows a considerable variation in nightly arrival times of both petrels, even unexplained when related to conditions of weather and light each night. However it will be noticed from these that the Leach's show fairly reasonable consistency in arriving, in general, about 20 minutes later than the storm petrels. Although the departure times are not given here for the reason stated, they are far more consistent than the arrival times, both birds apparently leaving regularly around 02.20 - 02.30hrs.
One day in bright sunshine a storm petrel left a burrow on the Holm (14.00hrs) and flew over two of us on its direct path to the sea; this is very unusual however.
Measurements and weights
A large number of the Leach's and storm petrels caught were weighed, and wing measurements were taken for about the same number of birds. The average weight of the storm petrels was 25.6gms. with a range 2o-33gms. It is interesting to notice that this is almost exactly the same as Waters found with his St.Kilda birds in June, although his were even lighter in Aug.(see table below). On the other hand both his and the Mykines birds are clearly lighter, albeit by a small amount, than other samples taken in S.W.Wales and Ireland, all of which show considerable consistency at around 28gms.
The female storm petrel that 1aid an egg whilst being handled, weighed 32gms before laying and 26gms after.
The average wing length of the storm petrels was 120.7mm. with the range from 108mm to 129mm. Again corelation can be found with the figures from S.W.Britain and St.Kilda, although this time the northern birds tend to be the longer winged ones.
The average weight of the Leach's petrels was 43.92 gms (range 32-58). As a point of accuracy however it should he remembered that a lot of our petrels regurgitated before weighing, so that in fact the average weight should be slightly higher if this factor is taken into account. Average wing length was 158.1 (147 -170mm.).
Table of comparative storm Petrel weights
Table of comparative storm petrel wing length
Size of sample Range Average( gms.) Mykinesholm B.E.G. Jul-Aug 834 20-33 25.6 St.Kilda E.waters June 27 22-29 25.27 Aug. 35 23.68 Skokholm F.L.Davis - ? ? 28.0 Gt.Skellig R.Lovegrove Aug 69 23-36 27.3 Inishtearaght R.Lovegrove June 81 23-35 28.2 Aug 70 23-33 27.7 Inishvickillane E.Byrne June 50 25-35 28.8
There appears to be no obvious correlation between weights and wing lengths at the extremes of the ranges heavier birds may have shorter, longer or about average wing lengths, and vice versa. Simon Boyd also plotted wings-and-weights against the state of development of the brood patch for both species but again no pattern was apparent.
Size of sample Range Average(mm.) Mykinesholm B.E.G. 847 10S-119 120.7 St. Kilda E.Waters 69 113-119 121.0 Skokholm P.E.Davis ? ? 117.5 Inishvickillane S..T.Byrne 50 110-123 117.4
The colour of the foreheads of storm petrels is variable.
Normally (i.e. the majority of birds) it is the same shade of dusky black as the remainder of the head, but sometimes it is very pale, and almost off-white in extreme cases. The area involved is also variable; sometimes the "mask" is large, up to the level of the eyes; sometimes it is very small, restricted to a cuticle-shaped band round the top of the bill.
The reason for this variation is not known. However one fact became apparent in relation to this feature and that is that the "pale faces" seem to be less numerous here than they are in the enormous S.W. Ireland colonies. In Ireland I found that on average 9.7% of the birds showed substantial pale faces (two years' figures) whereas on Mykines in a sample of about 850, only 3.7% showed the same feature, and of these over half were only slightly developed. Whether this feature is related to age or is some form of plumage variation, or what, remains unknown.
A measurable percentage of both species of small petrels has some degree of leg- or foot injury. It is quite frequent to find a leg completely missing above or below the ankle joint. Similarly, the tarsus is sometimes broken and reset often at a grotesque angle. Again, the foot may be missing or the toes abbreviated or missing.
The cause of these injuries is unknown but one can imagine that the considerable probability is that they are due to the birds pattering action when feeding on the surface of the sea and are caused by fish. If this hypothesis is correct one might wonder how much mortality is caused in this way. The injuries do not appear to hamper the birds in feeding, for all appeared to be in good health and the average weight of the 28 stormies (25.5gm.) and the 12 Leach'S (43.4) are both the same as the overall averages.
Leach's petrels 12 out or 400 = 3% with leg injury Storm petrels 28 out of 847 = 2.3% with leg injury
Brood patches Every one of the 1200+ petrels which we examined (Leach'S and storm) had a brood patch in some state of development. Whereas in most species the possession of a brood patch indicates breeding, with these two petrels this is not the case and all the birds appear to have a brood patch during the breeding season. We recorded the state of the brood patch in four degrees of development. These were
The answer to the problem of the breeding status of these two species presumably lies partly in the understanding of these brood patches and a knowledge of the pattern or regrowth of the down could very well throw light on the matter.
(a) vascularised i.e. fully developed with blood vessels dilated- in other words birds almost certainly incubating eggs, (b) fully developed (but unvascularised), (c) down present on the brood patches, in some degree but not fully covering the patch, (d) completely down covered.
In the Leach's petrels only 4% overall had vascularised brood patches, 40% fully developed (unvascularised), 61% partly downcovered, and 11% had completely down covered patches. The meaning of these figures broadly supports the opinions stated earlier as to the general stage of the breeding season at the time of our visit. i.e. few birds still incubating (4% vascularised), the majority feeding young (ages of young can be represented by the three degrees of regrowth of down - 40% adults had fully developed but unvascularised brood patches (newly hatched young?) 61% had down regrowing and 11% had down completely covering the brood patch (oldest young). However the simplicity of this thesis is spoilt by the problem of haw and where the non-breeding population fits into this pattern.
With the storm petrels, the percentages were as follows: vascularised 9%; fully developed 47%; some down 35%; complete down 9%. Here the suggestion is that the 9% vascularised birds represent the birds at the beginning of the incubating stage, while the 47% with fully developed patches must include many of those birds Shortly to lay eggs. A non-breeding population also has to be fitted in somewhere in this pattern, but how many and where is again unknown.
It will be remembered that the storm petrels were thought to be a few weeks later breeding than the larger species.
At advanced stages of the breeding season Leach's and storm petrels regurgitate food readily (not to be confused with the emission of oil which is a usual nervous reaction of the bird when handled). The following are the numbers of both species which were recorded as regurgitating (given as a fraction of the total recorded):
The actual figure for Leach's is in fact higher even than this, because it is known that some birds regurgitated when in the nets and the fact was not always passed on to the recorders. One very consistent fact emerges on analysis; every Leach's petrel recorded as regurgitating had a brood patch covered with some degree of down - as opposed to vascularised or fully developed (clear of down}.
Storm petrel Leach's petrel 6/1881 88/400
This presumably illustrates again the number of Leach's which had reached the chick stage, and conversely the very few stormies that were at an advanced stage of breeding.
We brought home regurgitated food samples from petrels which have been analysed as follows:
1. Petrel 25/7/68 Fish remains 100% 2. Leach's Petrel 25/7/68 Ditto, incl. otoliths and eye lenses 80%
Shrimp passiphaea 20%
3. Leach's Petrel 26/7/68 Fish 100% (Gadiculus thori, one; Gadus sp. one nematode) 4. Storm Petrel 27/7/68 Fish remains 70% (incl. eye lenses from 7 fish)
Shrimp hymenodora gracialis, one, 25%,
One Euphauriidd sp.
One Hypernid amphipod-themisto sp.
5. Storm Petrel 27/7/68 Fish remains 70%
One shrimp passiphaea sp. 15%
One Euphauriid, Meganychyhanes norvegica 10%
Two Copepod, Pareuchaeta norvegica
6. Fulmar 27/7/68 Fish remains 100% (apparently not trawler waste)
I am very grateful to Dr. Ivor Rees, Menai Bridge, for having carried out these analyses. He comments that a lot of the fish appear to have been about 7 cms. long; quite 8 mouthful for a small petrel. Also of interest is the fact that both species of shrimp recorded Passiphaea and Hymenodora are deep living crustaceans, but had presumably been taken at the surface. The fish making up the fulmar sample was similar to the fish in the small petrel samples.