Wikipedia talk:WikiProject Scottish Islands/Islands by area

Scalpay

 * 1) 66 = which Scalpay? --MacRusgail 09:44, 22 August 2007 (UTC)


 * Scalpay, Outer Hebrides - please note that this list is still incomplete. A more complete list is at list of islands of Scotland which should have greater prominence in the project. Finavon 19:14, 22 August 2007 (UTC)


 * I've been using both on and off, but find this address easier to access via WPSI. --MacRusgail 11:32, 23 August 2007 (UTC)

December 07 update
Initial finding - Eilean Ruairidh Mòr, 	Inchfad, and Eilean Sùbhainn (Loch Maree), are all missing from the main List of islands of Scotland.
 * Inchfad was already in the list; Eilean Sùbhainn - page created and added to list; I will investigate Eilean Ruairidh Mòr (also in Loch Maree I believe). The following group are not corrected - I'm only at Wiay, Skye tonight and am working down the list. Finavon (talk) 01:36, 14 December 2007 (UTC)

Sorry about Inchfad - I'd forgotten the 'c' and the sorting didn't have it under '40'. Ben MacDuiTalk /  Walk  19:30, 14 December 2007 (UTC)

A quick sweep of the infoboxes - with the missing Eilean Sùbhainn included I get Ben MacDuiTalk /  Walk  22:52, 13 December 2007 (UTC)
 * Hunda - 147
 * Ailsa Craig - 148
 * Soay, St Kilda - 149
 * Lunga, Treshnish Isles 159
 * Eileach an Naoimh 184
 * Horse Island - 188
 * Moncrieffe Island 199


 * I'm also not convinced that Na h-Eileanan Iasgaich are 50ha - that's presumably only at low tide. They have a plural Gàidhlig name, which is a giveaway --MacRusgail (talk) 23:19, 13 December 2007 (UTC) p.s. Is there anywhere I can access a list of islands from 30-40ha? I'm not sure all are noteworthy, but some will be.


 * We've been going with Hamish-Smith, except where there is a clearly better source. I suppose we need to decide if they stay as a single page or get one for each island. I think there will be so little information (and certainly little that differs within the group) that we should leave them as a group. Finavon (talk) 01:36, 14 December 2007 (UTC)


 * This is clearly one of those 'islands' that drove H-S to his overly prescriptive definition. My interpretation of our definition is that the individual bits are, at certain stages of the tide, individual isles, but that so is Eileanan Iasgaich itself. Its not alone - see Lunga, Firth of Lorn.

Re Ceann Iar (how is this pronounced differently from 'Ceann Ear' one wonders?) - I see this has just been amended from 154 ha to 134 ha, but my H-Smith has 154 (page 206).


 * Same page in 1997 reprint (so probably older) has 134 ha. I'm happy to go with the later edition, but would like another source as well. Finavon (talk) 21:09, 14 December 2007 (UTC)


 * Agreed - none to hand tho' I had a go at measuring it and unhelpfully came up with the answer 144 ha. Of course, at this size it depends on the assumed state of the tide in any case. Ben MacDuiTalk /  Walk


 * Remeasured at 121ha on OS 1:50,000 map Finavon (talk) 20:46, 16 September 2013 (UTC)

No idea where a list of smaller islands might reside. As they get smaller the definitional problems simply increase. Ben MacDuiTalk /  Walk  19:17, 14 December 2007 (UTC)


 * Kyown Eer as opposed to Kyown Air. "Ow" as in "Cow". The Monach Isles like the Outer Hebrides from Harris down are mostly sandy blobs, which makes their area hard to define. --MacRusgail (talk) 20:11, 14 December 2007 (UTC)

Where is #177 Garbh Eilean? - not Shiants as linked and not in H-S list. My suggestion is Loch Maree. I have started Garbh Eilean (disambiguation). Finavon (talk) 21:09, 14 December 2007 (UTC)

I am a bit sloppy with the links I'm afraid and have come to rely on your forensic skills in that department. 'Garbh Eileans'? - you've seen one, you've seen 'em all. You are of course right about Loch Maree - see Wikipedia talk:WikiProject Scottish Islands/Archive 1. Ben MacDuiTalk /  Walk  12:19, 15 December 2007 (UTC) PS I once visited the University of Leeds - the science building there claimed to have the 'longest corridor in Europe'. They started building it at both ends, and it didn't quite meet in the middle, so the corridor had this odd little chicane in it. Let's hope we don't have the same problem sorting out this list!

No chicane - well done sir! I was afraid of discovering a stray Garbh Eilean in our midst. Ben MacDuiTalk /  Walk  21:31, 16 December 2007 (UTC)


 * I thought more like tunnelling! I still aim to provide a skeleton for each island on the main list - down to Fuiay which does not appear to have a page yet. Finavon (talk) 22:36, 16 December 2007 (UTC)

Other "problems"
I have being looking over my OS Atlas of Great Britain, and trying to find problem islands. Oldany looks pretty large to me, but isn't in H-S as it is tidal. Eilean Mhic Chrion, which is off Eilean Righ looks similar in size to it, and Island Macaskin, but I suspect it too is tidal. Innis Mhòr off Easter Ross, probably fluctuates in size frequently, but falls into the "boring sandy blob" category, I think! I also want to find out which is the island in a loch in an island in Loch Maree. I find the idea of an island within an island within an island strangely intriguing, and kind of "Zen". --MacRusgail (talk) 17:10, 15 December 2007 (UTC)
 * 1) Oldany Island is already in;
 * 2) Eilean Mhic Chrion looks valid (?40ha);
 * I make it about 44ha. Good one. (All I'm doing here is taking a rough measurement, it's not anything scientific). Ben MacDuiTalk /  Walk


 * 1) Innis Mhòr looks like a sandbank. I presume an island needs to have something above the surface at high water. Otherwise it is a "drying" rock, reef etc. However if it is in, then there is another, unnamed dune just to its south east!
 * See Wikipedia talk:WikiProject Scottish Islands/Archive 1. I measured it at a fraction under 40 ha and put it on the smaller island list. Innis Beag is the one to the south east. Ben MacDuiTalk /  Walk
 * I don't know about Loch Maree islands, but West Loch on Hildasay has an islet. Finavon (talk) 10:12, 16 December 2007 (UTC)
 * The 'Wishing Stone' image on Erraid would seem to be a micro example. Ben MacDuiTalk /  Walk

There's also several islands near Forres in the River Findhorn estuary (probably boring sandy blob category) and at least two substantial ones in the River Conon in Easter Ross. --MacRusgail (talk) 17:34, 16 December 2007 (UTC)


 * Findhorn Bay definitely has no islands at high tide. If the ones in the Conon are Garrie Island and the blob to its north east, I'd say they are under 40 ha. Ben MacDuiTalk /  Walk  21:46, 16 December 2007 (UTC)
 * I see:
 * Dunglass Island at Conon Bridge ?40ha max 8m - channel re-opened by recent work;
 * Eilean Aigas on River Beauly 60 acre (24ha) ?117m - in Castle island list;
 * Beauly ?12ha ?13m - getting a bit small;
 * also Merkinch (Inverness) ?350ha is formed by river and canal (so not an island!). Finavon (talk) 22:43, 16 December 2007 (UTC)


 * Dunglass is about 40 ha. You could try this freebie at: . It would not seem to be Mac friendly so I can't use it. Ben MacDuiTalk /  Walk  08:50, 17 December 2007 (UTC)


 * The island by Beauly is marked as wooded, which suggests it is not just some temporary phenomenon. Interestingly, Merkinch contains "-innis", as suffix, which suggests it may have been an old island (or at least in a marsh). Others - Moy Island (Conon NH487545 - but small), ? (Black Water/Conon tributary by Contin, NH457566, more substantial), Eilean an Daraich (NH439590)--MacRusgail (talk) 15:20, 6 January 2008 (UTC) p.s. Eilean Hoan (NC443674) listed anywhere?


 * The first group all look like 'permanent' (by human standards at least) islands, although I don't think any reach the magic 40 ha. I am looking into Eilean Hoan. Ben MacDuiTalk /  Walk  16:57, 8 January 2008 (UTC)

Findhorn delta islands (sandy blob category) - There is also a substantial tract of land which doesn't appear as an island on the current map, but has on some others. This is to the north east of Moy House. All of these are best viewed at maximum magnification. Getamaplink --MacRusgail (talk) 15:06, 6 January 2008 (UTC)
 * NJ022625
 * NJ028619 (quite large?)
 * NJ054616

If you look at 'em on Google maps, you can see that they are saltmarsh 'islands' and that the sand is only visible at low tide. There is no question that they are islands but they lack permanence as the rivers change their courses. I don't think any are at 40 ha, and none seem to have names. Ben MacDuiTalk /  Walk  16:50, 8 January 2008 (UTC)


 * Sure, I tend to agree. But if they are wooded, that suggests something more permanent perhaps. After all, the Isle of Sheppey is a sandy blob really and has a long and involved history. --MacRusgail (talk) 19:39, 8 January 2008 (UTC)

Eilean Hoan
According to the JNCC the SSSI it is part of, which includes Eilean nan Ron and other neighbours is 221.11 ha. As EnR is only 138 ha I wondered if H-Smith might have missed it. However, as a rough estimate I make it about 26 ha. H-Smith mentions its existence under "Eilean nan Ron", so it is a good find for the 'Smaller' list. (For future ref the Latitude & Longitude are N58:34:01 W04:39:39 (Lat: 58.567 Long: -4.661)). Ben MacDuiTalk /  Walk  17:58, 8 January 2008 (UTC)


 * There's also Eilean Horrisdale NG787744 - possibly tidal, with some evidence of settlement (Cnoc an t-Sean Bhaile meaning old town hill and Tigh Ghobain - another glorious OS bastardisation which probably is an old Smithy). If you're looking for it, it's near Badachro on Loch Gairloch [sic] (surely it should be just "Gairloch"?!) --MacRusgail (talk) 19:36, 8 January 2008 (UTC) p.s. 38m at highest point, and seemingly fairly rocky.

Eilean Heast
This came to my attention courtesy of Heaste. Its a tidal islet of about 20 ha. Ben  Mac  Dui  19:56, 30 January 2009 (UTC)

Ooh yeah
Uyea, Northmavine, Shetland is not on the list, but apparently "125 hectares" --MacRusgail (talk) 13:53, 8 January 2008 (UTC)

H-Smith says it is 45 ha - no idea where that came from. I'll fix it and throw in a free infobox if I have time. Ben MacDuiTalk /  Walk  14:47, 8 January 2008 (UTC) PS I had no idea you were a fan.


 * I'm more into An Orkney in the Uyea myself. Although as far as prog rock goes, I like some of Pink Floyd's early stuff... --MacRusgail (talk) 19:37, 8 January 2008 (UTC)

Infobox refs
The GA review of Gigha has raised a number of issues, including the use of references in the infobox (see Talk:Gigha). The solution was to use a new format that involves a main reference stating "Infobox source is Haswell-Smith (2004) pp. 37–41 unless otherwise stated." and in-line refs for others. This means using something like "Haswell-Smith (2004) pp. 502–03. Modified to include bridged islands." for the area rank. It occurs to me that we could identify in the table which those are. Ben  Mac  Dui  19:46, 30 January 2009 (UTC)


 * I'm beginning to think that part of the problem of Wikipedia is its instability. People keep on changing the goal posts. Usually in places which few others know about... who came up with this new format? --MacRusgail (talk) 18:11, 31 January 2009 (UTC)
 * Pass. You can use Template:Cent to keep track of "central discussions", but there are a lot of them and part of the problem is that some of them recur, with the same arguments repeated ad nauseam. Ben   Mac  Dui  18:28, 31 January 2009 (UTC)

Boreray, St Kilda
Following from a discussion on the talk page its area has been upped to 86 ha. The following infoboxes needed to be amended. Ben  Mac  Dui  15:41, 1 September 2009 (UTC) ✅


 * 153=
 * Boreray, St Kilda


 * 153=
 * Eilean Rìgh


 * 155=
 * Auskerry


 * 155=
 * Eilean Chaluim Chille


 * 155=
 * Oronsay, Outer Hebrides*


 * 158=
 * Fuiay


 * 158=
 * Vuia Mòr


 * 160
 * Lunga, Treshnish Isles


 * 161
 * Balta, Shetland


 * 162=
 * Muldoanich


 * 162=
 * Bigga


 * 164=
 * Eilean Kearstay


 * 164=
 * Inchlonaig

Stromay
This is now included at 66 ha. The following infoboxes need to be adjusted. Ben  Mac  Dui  19:17, 1 September 2009 (UTC)

From 174=, Stromay downwards.

All ✅. Ben  Mac  Dui  19:59, 12 September 2009 (UTC)

Dunglass
Dunglass Island arrives - estimated at 55 ha but good to have a 2nd opinion before adjusting the lists. Ben  Mac  Dui  11:19, 24 January 2010 (UTC)

Calbha Mor
A tidal island in Enard Bay at. Approx 50 ha and second opinion requested. Ben  Mac  Dui  12:45, 27 November 2011 (UTC)
 * Agree. OS show Calbha Mór as tidal, with an area of 61ha. Its smaller neighbour, Calbha Beag is 31ha. See also and PDF Finavon (talk) 20:03, 11 December 2011 (UTC)
 * Many thanks. Ben   Mac  Dui  09:00, 12 December 2011 (UTC) PS - Very useful pdf, which I haven't seen before.  Ben   Mac  Dui  09:02, 12 December 2011 (UTC)

New Tables
The Rick Livingstone tables (above thread) are producing new issues. Some are straightforward in that he provides an estimate that is different from the existing OS estimates. For ranking purposes the current logic is:
 * Haswell-Smith is the most reliable source.
 * Gazeteer for Scotland and GRO come next as established refs.
 * Livingstone's tables are next in rank.
 * Our estimates are last in line unless otherwise discussed.

Perhaps ideally the table would include all the estimates, but that's a task for another day, for me at least.

The only one that I am wondering about so far is Stromay. His estimate of 100 ha is much bigger than our 66 and I suspect includes a lot of rocks covered with water at low tide. Not sure what to do yet.
 * OS shows Stromay as 65 or 66ha along the high water mark, but double that at low water (ie drying rocks). Finavon (talk) 13:29, 18 December 2011 (UTC)

The second issue is that he has "found" more islands:
 * Ceallasaigh Beag at a whopping 55 ha. I also not that at low tides it is connected to Ceallasaigh Mor, making it 110 ha - an anomaly that can be explained in the text of the articles I think.
 * closer to 46ha Finavon (talk) 13:29, 18 December 2011 (UTC)


 * Calbha Beag at 45 ha.
 * 31ha at HW Finavon (talk) 13:29, 18 December 2011 (UTC)

There may be others Ben   Mac  Dui  12:13, 17 December 2011 (UTC)
 * At a first look, Livingstone also lists the following, which are not in the current area list:


 * Gluss Isle - promontory
 * St Ninian's Isle 1.3 km2 - promontory
 * Fora Ness - promontory
 * Eilean Meadhonach (Crowlins) 77 ha
 * Eilean an Taighe 55 ha - joined to Garbh Eilean
 * Eilean Dubh Mòr 50 ha
 * Flodday (Loch Maddy) 50 ha
 * Eileanan Chearabhaigh 49 ha - group
 * Nave Island 40 ha

It becomes more confusing below 40 ha! Finavon (talk) 13:29, 18 December 2011 (UTC)

No kidding. I'd much prefer a genuine RS to our estimates, but some of Livingstone's do seem questionable. Of the above that are "islands":
 * Eilean Meadhonich is part of the overall area of of Eilean Mòr, Crowlin Islands in the H-Smith system but I do think we should add it and amend Eilean Mòr.
 * Eilean an Taighe isn't normally separated from Garbh Eilean and should not be added.
 * Eilean Dubh Mòr is in with the H-Smith estimate of 65 ha
 * Eileanan Chearabhaigh is a complicated series of islands and non of our estimates have  shown any of them to be >40ha at high tide  and should not be added.
 * Flodday (Loch Maddy) - mentioned by H-Smith, not tidal and therefore <40ha in his system.
 * Nave Island - mentioned by H-Smith, not tidal and therefore <40ha in his system.

Good work on the new maps btw. Ben  Mac  Dui  18:39, 18 December 2011 (UTC)


 * If you are happy with the above i.e. adding Eilean Meadhonach, Crowlin Islands and the other revisions already done, I'll add this to the table and with any luck that's it. Everything under 137 ha will then need to be checked for the revised ranking. Ben   Mac  Dui  19:12, 18 December 2011 (UTC)


 * PS I have left Stromay at 66 ha. I presume you want me to downgrade Calbha Beag from the above. Ben   Mac  Dui  20:11, 18 December 2011 (UTC)


 * Done. I fear we may end up with several different lists if we are not careful. However it is still consistent with H-Smith in that he has 162, we have added a further 60 and there are 222 on the list. Ben   Mac  Dui  15:57, 28 December 2011 (UTC)
 * Well done - I missed the above reply. I think Calbha Beag is fine - I'm sure using the figure that appears to relate to HW is right. It is somewhat reassuring that the current table is supported by 2 independent sources. Finavon (talk) 17:30, 28 December 2011 (UTC)
 * Thanks - it would be good if someone would publish something definitive for freshwater islands, although on the other hand I'd prefer not to have to amend things again for a while at least. Note to self: File:Loch Midhinis - geograph.org.uk - 854594.jpg shows part of Ceallasaigh Beag.  Ben   Mac  Dui  18:34, 28 December 2011 (UTC)

PS There is another "complex island" Cliasaigh Mor and outliers just south east of the Ceallasaighs. It fits H-Smith's definition and he does mention it, so he must have measured it at <40ha. Ben  Mac  Dui  18:42, 28 December 2011 (UTC)

Tidal isles
According to edits made quoting Caton, Peter (2011) No Boat Required - Exploring Tidal Islands. Matador, the are 17 tidal islands in Scotland accessible from the mainland. There are numerous other tiddlers and I wonder what the lower size limit is. Ben  Mac  Dui  21:15, 10 January 2012 (UTC)
 * Hestan, Davaar, Eilean Ighe, Rough Island, Castle Tioram, Cramond, Eilena Shona, Mhic Chrion, and two Islands of Fleet have been edited. That's 10.
 * Danna, Eriska, Oldany, Calbha Mor, Eilean Tioram in Loch Gairloch?, Innis Mhòr, Eilean Donan, Eyebroughy - that's 8 more.

Update from Livingstone's Tables
I've completed a first pass at this. My main aim was to ensure than main lists were not missing anything - happily they weren't although there were a few Orkney Holms that needed 'promotion' onto the main table. It has also thrown up various issues.


 * Red links - only Lingeigh, North Uist above 40 ha. ✅
 * Of those between 16 and 40 ha the following is an approx listing by archipelago:
 * Shetland - 2 - Little Holm and one of the Lingas;
 * Orkney - 1 - Start Point is a redirect;
 * Outer Hebrides - 18
 * Inner Hebrides - 11
 * North Highland - 11
 * Mainland sea lochs - 4
 * Freshwater - 2
 * There's an interesting thesis on the psychological geography of the Scottish islands here for all our interns to get stuck into.


 * Anomalies for islands >40 ha, which I will come back to.


 * In theory, given Livingstone's claim that the tables are a "register of all islands 15 hectares, or more" it would be possible to extend the ranking right down to 15 ha. Potentially tricky and frustrating and not something I'm going to attempt any time soon.


 * There are quite a few of the smaller islands, especially in the Northern Isles, that lack an infobox.  Ben   Mac  Dui  15:56, 15 September 2013 (UTC)

Ranking anomalies

 * Eilean Trodday estimated from OS at 42, but our friend Rick has this at 38 ha, so it drops out of the 40 ha + rankings.
 * Re-measured at 42ha


 * Eileanan Chearabhaigh - I have included this for now at 49 ha. No idea how he gets this number.
 * Measured from OS at 32ha


 * Lingeigh, North Uist as above at 42 ha.
 * Agree 42ha


 * Stromay - RL has this at 100 ha.
 * OS shows 66ha (including the 2ha lochan)


 * Eriska
 * 112ha


 * Rabbit Islands - included by RL at a total of 44 ha
 * Nave Island - in or out?
 * between 40 and 41ha


 * Uyea Northmavine in at only 28 ha
 * measured at 41ha


 * Eilean Meadhonich, Crowlin Islands
 * Dunglass
 * Flodday (Loch Maddy)
 * Oldany
 * Eilean Mhic Chrion*
 * Ceallasaigh Beag
 * Areas from 1:25,000 OS maps on MHWS level, but I'm not sure if that helps as OR in the face of a published source! It might help select where several conflicting sources exist. Finavon (talk) 19:25, 16 September 2013 (UTC)
 * Thanks for this. (I wondered if I was going batty at Stromay.) It might be worth a check at the NOR noticeboard when there is a clar list of anomalies.
 * Will strike Uyea from the list below.
 * Do you want to argue (per 4 below) that at c 120 ha the Eriska Hotel trumps RL?
 * I'm fine with Stromay at 66 ha per #11 below for now at least.
 * Nave would potentially fall foul of #6 - although feel free to amend these suggestions.
 * Similarly, Trodday is a bore - the difference is small but its in a crucial area.
 * Chearabhaigh is probably better off the list anyway. Did you measure the whole thing or just bits? Ben   Mac  Dui  20:07, 16 September 2013 (UTC)

Less than 40 ha

 * Complex islands - no mention of Cliasaigh Mor in Loch Maddy or Maithidh Glas/Maithidh Riabhach and Maragaidh Beag/Maragaidh Mòr south of Ronay and east of Benbecula - the first and last of which at least look bigger than 15 ha at low tide.
 * Fuidheigh Beag and Orasaigh ditto - both also east of Benbecula. Ben   Mac  Dui  16:32, 15 September 2013 (UTC)

Attempt at an explicit logic
Definition of an island: "land that is surrounded by seawater on a daily basis, but not necessarily at all stages of the tide, excluding human devices such as bridges and causeways
 * 1) For ranking purposes, for islands >40 ha the area is per H-Smith unless no other estimate is available as this provides a consistency of approach.
 * 2) The next best reference is GRO for inhabited islands.
 * 3) The next best reference is Gazetteer of Scotland, which has academic credentials, followed by Livingstone.
 * 4) Failing that any other available estimate from a credible external source. If sufficiently reliable it may trump Livingstone.
 * 5) Failing that, an OS estimate.
 * 6) Islands may be included in the rankings for >40 ha if there is a reliable source for this, even if that it is clear H-Smith is aware of the island  unless there is some very clear evidence of an error or omission on the former's part. Similarly, if a reliable source exists for the area of an island that Haswell-Smith counts as a multiple island - then what?
 * 7) Bridged and tidal islands are included provided they meet the standard "land that is surrounded by seawater on a daily basis, but not necessarily at all stages of the tide, excluding human devices such as bridges and causeways" definition.
 * 8) A corollary is that although tidal islands are included their area has to be measured at a stage when they are "surrounded by seawater".
 * 9) 'Complex islands' are only included if measured by an external source as there is too much room for errors on OS estimates. Maybe they need their own mini table.
 * 10) The same would apply to islands <40 ha except that very few are measured by H-Smith.
 * 11) There needs to be some wiggle room here for common sense as Livingstone has made at least one error - see note to Eilean Horrisdale on project page.
 * 12) If there is doubt as to whether or not the object is an island, the Ordnance Survey is the principal reference for determining the extent of high tides.
 * 13) An island may only appear on the list once. Ben   Mac  Dui  18:29, 18 September 2013 (UTC)

Specific ranking issues

 * Oldany - contradicts above unless we place Gazetteer higher than RL.
 * Eilean Sùbhainn - ditto
 * Eriska - the issue here is a mix up between ha and acres (see ref on project page). It is only c. 121 ha according to the hotel website, so RL's 145 ha is fine. Definitely needs fixing.
 * Torsa - needs minor adjust to RL's 110 ha as this is an OS est at present I think.
 * Hunda - ditto
 * Boreray, St Kilda - is OK per No. 4
 * Stromay - RL has this at 100 ha but it looks no bigger than Hermetray both he and H-Smith have at 72 ha. An IAR per No 11?
 * Nave Island - in per No 6
 * Eilean Trodday estimated from OS at 42, but RL has this at 38 ha, so it drops out of the 40 ha + rankings.
 * Lingeigh, North Uist is in at 42 ha
 * Rabbit Islands - two at <40 ha per No 8
 * Eilean Meadhonich is in but needs a note on the main table
 * Eilean Mhic Chrion should drop to 40 ha per RL
 * Eileanan Chearabhaigh - I think that we can accept that this is an island of 49 ha at some stages of the tide per No. 6 although it is a real oddity. Listed at 32 ha per OS - can't be 49 ha per #6.
 * Ceallasaigh Beag - should be in at 55 ha, possibly with explanatory note.
 * Calbha Beag is in unless we can exclude per 11.
 * Uyea Northmavine is in although RL has only 28 ha.
 * Flodday, North Uist (Loch Maddy) is estimated by RL at 50 ha and is in.
 * Eilean Dubh Mor problem (below)


 * Innis Mhòr is in the <40 ha list. Can't find a photo of it. IN per No 5
 * Inchkeith - presumably RL thinks its less than 15 ha, but it can stay in the smaller list.
 * Fuidheigh Beag east of Benbecula - needs an OS estimate for smaller table.
 * Orasaigh - ditto, ditto.

Freshwater islands not mentioned by RL: ✅
 * Eilean Mòr, Loch Langavat
 * Rainish Eilean Mòr
 * Moncreiffe
 * Dunglass

The last two are river islands and RL may not have considered them - you can't sail round them after all. The Outer Hebridean lochs are hard going and its easy to miss something. I can't see any reason to exclude them.

Add to unranked 'complex islands' minitable (and main list of islands) ✅
 * Crowlin Islands
 * Cliasaigh Mor
 * Maithidh Glas/Maithidh Riabhach
 * Maragaidh Beag/Maragaidh Mòr
 * Rabbit Islands?
 * Eileanan Chearabhaigh

Subjective comment. Haswell-Smith gives the impression of having anchored at every location and spent some time with his measurements. Livinsgtone's tables are very helpful but I get the impression the method of measuring is less consistent. Perhaps a 'sail-by' estimate that depended on the state of the tide at the time? This might explain e.g. Gluss Isle, which he may have passed by during a high spring tide. Ben  Mac  Dui  12:00, 16 September 2013 (UTC)

I have only just noticed the reason why H-Smith has 65 ha for Eilean Dubh Mor and Livingstone has 50. H-Smith is including E Dubh Beag in the total as they are joined at some low tides. This is quite explicit on p. 72. Amended 'rules' above as a result but failed to come up with a satisfactory answer. The sad fact is that Livinsgtone's tables use a logic that is much closer to out own and attempting to use Haswell-Smith as a basis for area ranking is beginning to look increasingly thin. Perhaps we should keep the area per H-Smith but rank them per Livingstone?. Ben  Mac  Dui  19:06, 18 September 2013 (UTC)

Revised logic
I have had a look at this and most of the above issues dissolve if the logic is:
 * a) definition of an island
 * b) individual islands areas are resolved on a case-by-case basis using the most reliable sources available.
 * c) ranks are per Livingstone's tables, modified as needed by (b)

This is far less clumsy and avoids:
 * the need for complex modification of H-Smiths rankings - I've put Livingstone into a s/sheet and the first 34 in rank are all exactly as our existing ranks and most of the remainder are fairly trivial differences.
 * the arbitrary cut off of 40 ha. We can still assign ranks to there and no further if we wish but there is no need to stop there and thus we don't have to worry so much about whether or not Nave Island is or isn't ranked.
 * some if not all of the OR issues.

I'll come up with a short list of non-trivial differences from Livingstone's list and a revised list to sort out asap. Ben  Mac  Dui  18:52, 19 September 2013 (UTC)

Draft proposal
My plan here is:
 * to leave this logic here for a while and if there are no objections/amendments put it on the project page in say a fortnight.
 * re-order the ranking based on any changes - I think they are all already explicit above.
 * add a note of some kind to the project page explaining any significant differences from RL's ranking and the proposed one. I'll draft something here soon.
 * leave the table for a further week or two before amending the rankings on the article pages.
 * consider adding a note to the Infobox.
 * consider turning Mr Livingstone's tables into a template - it is clearly getting ever-wider use. Ben   Mac  Dui  07:49, 20 September 2013 (UTC)

Definitions
 * An island is: "land that is surrounded by water on a daily basis, but not necessarily at all stages of the tide, excluding human devices such as bridges and causeways”.
 * Bridged, tidal, loch and river islands are therefore included provided they meet the above definition.
 * A corollary is that although tidal islands are included their area has to be measured at a stage when they are "surrounded by seawater".
 * If there is doubt as to whether or not the object is an island, the Ordnance Survey is the principal reference for determining the extent of high tides.

Protocol for area measurement
 * 1) Articles use the best reliable source available on a case-by-case basis.
 * 2) For larger islands this is often Haswell-Smith (2004) and ideally this source is used for consistency.
 * 3) For inhabited islands the next best reference is often General Register Office for Scotland (November 2003) Scotland's Census 2001 table.
 * 4) Other useful sources are the ‘’Gazetteer of Scotland’’, which has academic credentials, and Rick Livingstone’s comprehensive tables.
 * 5) Occasionally another estimate from a credible external source may trump the others.
 * 6) From Ordnance Survey maps map be used in the absence of other sources or if it is clear that a source includes an error.

Protocol for ranking


 * 1) Ranking is based primarily on Rick Livingstone’s tables, but may be adjusted for amendments to measurements agreed above.
 * 2) 'Complex islands': an island may only appear on the list once and usually the smaller islands that appear at higher stages of the tide are used for ranking purposes.

Comments on RL's tables

 * Ulva (35) is estimated by HS at 1990 cf 19km2 by RL (the latter does not go for lots of decimal places for larger islands)
 * Eilean Shona (72) is estimated by HS at 525 cf 6km2 by RL
 * Danna (83) is estimated by Gaz at 315 cf 2.5km2 by RL
 * An implication of the above definition is that Lunga, Firth of Lorn's area does not include its high tide outliers. However the difference between H- Smith for the whole (254) and RL for the part (250) is not great and it only changes the ranking by 1.
 * Isle Ristol (103) is estimated by Gaz at 225 cf 1.5km2 by RL
 * Longa/Horrisdale RL error referred to above.
 * Papa Stronsay (72) is estimated by HS at 74 cf 54 by RL

RL has a total of 347 islands.
 * Losing St Ninian’s Isle, Fora Ness, En an Tighe Shiants & Gluss Isle as not islands plus GB as irrelevant gives 342.
 * Adding back Dunglass, Moncreiffe and Eilean Mor Lonagavat gives 345. It all adds up! Ben   Mac  Dui  08:22, 20 September 2013 (UTC)

Unresolved issues
There may be queries about individual islands where OS estimates differ from others but the only two missing any estimate are:
 * Fuidheigh Beag east of Benbecula
 * Orasaigh, Benbecula - ditto

Plus
 * GRO have Flodda at 145 ha and RL has 70. The latter may be more accurate?
 * GRO have Grimsay SEB at 117 ha and RL has 68. The latter may be more accurate?
 * GRO have Davaar at 54 ha and RL has 100. The former may be more accurate?
 * Stromay and Eilean Mhic C seem to be better estimated from OS than RL
 * I have listed Uyea Northmavine at 28 ha per RL but I think this should revert to HS at 45 ha.
 * I'm havering over Stromay. Superficially it does look as if it just about fits into a one km square on the OS maps.

There are numerous other smaller discrepancies but few of any substance. Ben  Mac  Dui  08:50, 24 September 2013 (UTC)