Earring Length: The Millimetre Gap Between Earlobe and Chin

The gap between earlobe and chin, in millimetres

The construction uses three published numbers and one declared here. Face length is the run from the mid-point of the hairline to the tip of the chin; cheekbone width is the widest horizontal measure across the zygomatic arches; the ratio of the first to the second is the lead check in this site's rule set. The proportion that rule set was fitted against divides the same run into three equal parts, and the lower part stops at the chin. A drop hangs from the ear, so the room it has before its lowest point passes the bottom of the face is that lower part — face length divided by three, which is the same as the ratio multiplied by cheekbone width and divided by three.

The number a listing prints is not this. An earring post length of 8 mm says how far a wire goes through an ear, and it reads the same whatever the face underneath it does. Two of the pages in that top-ten set repeat the figures 5, 6.5, 8 and 10 mm, and one of them prints a post ladder of 5, 6, 8, 10 and 12 mm while stating in the column header that the charm and the backing ball are excluded from it. Nothing in that set measures a drop against a face, so nothing in it tells you which of its own numbers was supposed to be compared with anything.

The table indexes the gap by ratio, because the ratio is the quantity published here for every outline. The three width columns are there because one ratio at 128.2 mm and the same ratio at 148.8 mm are different faces carrying the same number, and the gap between those two columns is 6.2 mm at the bottom row of the ladder and 12.4 mm at the top.

Every cell is one multiplication and one division — ratio × cheekbone width ÷ 3 — carried out exactly and then rounded to one decimal, so any cell can be re-derived on the spot. The inch column is the millimetre cell divided by 25.4 (40.8 ÷ 25.4 = 1.6063, so 1.61), and the last two columns add one inch to the 136 mm gap, which is how far a drop must reach to sit 25.4 mm below the chin at that width. The three widths are the adult cheekbone band this site works with: 128.2 and 148.8 mm are one standard deviation either side of the measured sex means, and 136 mm is the working line used by two of its eyewear pages. The ratio band is 1.05 to 1.65 with a published standard deviation of 0.18, and the ladder runs 0.90 to 1.80, which is 0.15 past each published end — 0.82 of that deviation rather than a whole one. One full deviation beyond the ends would be 0.87, a gap of 39.4 mm, and 1.83, a gap of 83.0 mm; each of those sits 1.4 mm outside the two ends of the ladder as printed. Two things are this page's own and are labelled as such. The first is the landmark. The lower part of the face runs from the base of the nose to the chin, and no position in the comparison below puts the earlobe at a measured level on the face, so this table takes the lobe to sit on that boundary; if it sits 6 mm higher or lower, every cell in the 136 mm column moves by 6 mm (0.24 in) and no row reorders. Six millimetres is just under the span of the widest hardware ladder named above, the post ladder that runs from 5 mm to 12 mm; that is why the sensitivity is stated this wide and not tightened for effect. The second is that the outer columns and the two end rows are combinations, not typical faces: the top row at the wide column implies a face length of 267.8 mm, which is more than a tape finds on anyone. Read one row at your own width, not the corners.

Where a printed number actually stops

Set the numbers a listing sells against that gap and the sizing vocabulary comes apart usefully. The next table subtracts the gap from the drop, so a negative cell means the lowest point is still above the chin and a positive one means it has passed the bottom of the face. Three ratios are used: the two published ends of the tape band and its middle, 1.05, 1.35 and 1.65, which give gaps of 47.6, 61.2 and 74.8 mm at the working width.

Two readings matter more than the shape of the table. A 30 mm drop stops above the chin on every row of the ladder, because the smallest gap in the previous table is 38.5 mm, so the short-drop band that a March 2026 guide defines as 20 to 30 mm cannot reach the bottom of any face in this set. A 50 mm drop, the figure that same guide uses as the floor of its long-dangle band, passes the chin on three of the ladder's ten rows at the working width and on two of ten at the wide end of the cheekbone band. Which side of the face a number falls on therefore depends on the face at least as much as on the number.

The second reading is the one the pages cannot supply, because it needs both columns at once. A shopper moving from the advertised short-drop band to the advertised long-drop band travels 25 mm, from roughly 25 to roughly 50. The room those bands ought to be judged against is 40.8 mm wide down the 136 mm column of the first table, from 40.8 mm at ratio 0.90 to 81.6 mm at ratio 1.80 — 1.63 times the distance the sizing vocabulary itself travels. Read corner to corner across all three cheekbone widths it is 50.8 mm, from the smallest cell of that table to its largest, which is exactly 2.00 in and 2.03 times that vocabulary. The categories are coarser than the room they are supposed to fit.

Rows are drop lengths as they appear in the pages this subject ranks for; columns are those rows minus the gap for that ratio, so the arithmetic is one subtraction per cell and every input is in the previous table. The percentage column divides the drop by the 1.35 gap of 61.2 mm, which makes 100% a drop ending level with the chin on a face at the middle of the published band; the inch column converts the signed millimetre cell by 25.4 and keeps its sign, so -0.44 in is eleven millimetres short of the chin, not eleven millimetres past it. Three rows are worth reading as crossings: 50 mm is the first printed figure to go positive at ratio 1.05, 70 mm is the first at 1.35, and 80 mm is the first at 1.65 — 60 mm still stops 1.2 mm short of the chin at that middle ratio. The figures that turn up most often in the hardware tables above are 6 mm and 8 mm, each printed by nine of the twelve pages read here, and 8 mm is 13.1% of the middle gap of this table. The longest post figure in that set, 12 mm, is 25.2% of the narrowest column here and 14.7% of the longest gap at the working width, which is the honest size of the biggest number a shopper is handed.

One third, and what the third was for

The division by three is not this page's invention, and it is older than anything else in the comparison below. A beauty and dress manual published in 1901 quotes a painter's definition of a pretty woman's face, and the clause it quotes is a construction: “A face of regular beauty should have the three following dimensions of equal size”. The three the book then names run from the summit of the forehead to the eyebrow, from the eyebrow to the end of the nose, and from the end of the nose to the extremity of the chin. That is the lower third this page divides by, stated one hundred and twenty-four years before the oldest row in the table below that puts a millimetre figure and a face word in one sentence, and one hundred and twenty-five before the newest dated row in it.

The same book has an opinion about earrings, and it is the opposite kind of opinion. The length of earrings is handled there as weight rather than as distance: the recommended thing is a small, light screw which “does not weigh down the lobe of the ear to an ungraceful extent”, and the enormous pendants are left to women of a certain age. Read beside the rest of the set, that is a fair summary of the whole comparison. The face was divided into measurable parts long before anybody measured an earring against those parts, and when earrings were discussed the discussion was about heft and propriety.

Two limits sit on the third itself. Three equal parts are an ideal rather than an anatomical finding, and the reason to measure a face at all is to find which of its three parts departs from the division; a face running long in its lower part has more room than the table gives it, and the table describes the even case. The midpoint of that lower part — the halfway stop, at 20.4 mm on the shortest row and 40.8 mm on the longest at the working width — is the only other position in this construction that comes out of the arithmetic alone and not out of an assumption.

What crossing the line costs, in the site's own units

Once a drop has passed the chin, its lowest point sits below the bottom of the face by a distance, and that distance has a size in the units this site already uses. Read as geometry, a drop ending x millimetres below the chin extends the silhouette from hairline to lowest point by x, which is x ÷ 136 of ratio at the working width. So 13.6 mm past the chin buys 0.10 of face-length ratio, and 0.10 is 0.55 of the ratio's published standard deviation. One millimetre past the chin is 0.007 of ratio, or 0.04 of a deviation.

Worked on a single face: at ratio 1.35 the gap is 61.2 mm, so a 70 mm drop ends 8.8 mm past the chin, which is 0.065 of ratio geometrically and 0.35 of a deviation. The only transfer of millimetres into apparent length published anywhere on this site is on beards and not on earrings, where 15 mm grown below the jaw was put at 8 to 10 mm of perceived face length, a transfer of 0.53 to 0.67. At that rate the same 8.8 mm reads as 4.7 to 5.9 mm, which is 0.035 to 0.043 of ratio.

That is the shape of the limit, and it is stated here as one. The geometry gives a ceiling. The only measured rate on this site belongs to hair attached to a jaw, not to an object hanging clear of the face, and nothing here has been calibrated against perception. So this page locates the crossing and declines to price it — which is enough to act on, because crossing the bottom of the face is a different event from stopping just short of it, and by how much a crossing reads is precisely the part nobody has measured.

The two measurements, and the trap in the third

If you have a tape, you do not need the ladder. Measure face length from the mid-point of the hairline to the tip of the chin and divide by three; that is the gap, and the ratio never enters the arithmetic. The ladder exists for the other case, where the ratio is the number you already have from a calculator or a shape guide, and it exists in three width columns because a ratio by itself does not fix a distance.

The trap is that two scales of this one ratio circulate on this site and they are not interchangeable. The detector reads a shorter face from the same photograph than a tape does from the same head, so the ratio it works in runs 1.10 to 1.26 while the tape scale every page here is written in runs 1.05 to 1.65; what the calculator shows you is the internal figure converted onto the tape scale. Feed the internal number into the ladder instead of the displayed one and the heart outline's centre becomes 1.182 instead of 1.35, which returns a gap of 53.6 mm where the published centre returns 61.2. That is 7.6 mm of gap lost — most of the distance between two adjacent sizes a listing sells.

One published thing to point at rather than reprint. The corridor arithmetic on this site's heart-shaped-face eyewear page is this same construction, ratio × 136 ÷ 3 with a declared allowance taken off it, spent on how deep a lens can be rather than how far a drop can fall. The two numbers come out of one division and belong to different objects, so the eyewear page's own figures are left where they are.

What has been printed about this, in date order

The rows below are the readings behind the first paragraph of this page. Positions are described by what they are and none of them is linked. Dates are given as each position carries them, including the two kinds a reader cannot see on the page itself. Quotations are verbatim, except the 1901 row, whose text was recovered by optical character recognition and is quoted only in the clause that survives that cleanly.

Two things changed across the span and one did not. Units arrived: the oldest row carries no unit of length for an earring at all, and seven of the eight dated web rows after it print millimetres — the one that does not is the 27 August 2024 row, which argues from face shapes and prints no measurement in any unit. The drop got a number late, and it got one in only two places, both of them outside the ten highest-ranked positions: a February 2026 guide prices long drops at 25mm–50mm for what they do to a neck, and a March 2026 guide puts short drops at 20 to 30 mm and longer dangles at 50 mm or more. What did not move in one hundred and twenty-five years is the join. No position in the set puts a millimetre figure, a word for a drop and a word for a face into one sentence. Six of the eight readable top-ten rows print millimetres, and all six spend them on the post, the gauge or a diameter. The few sentences elsewhere where a millimetre and a face word share a line are about where a forward-helix piece sits against the head, about measuring a face in millimetres and then spending the answer on a stud held against an earlobe, and about hoop diameters for longer faces on a page ranked eighteenth.

Sorted by publication date, with the four positions carrying no date placed last. Dates are recorded as each position itself carries them: printed where a reader can see them, printed in structured data only, or absent — the two structured-data-only rows and the four undated rows are stated rather than smoothed over, because they are the difference between a page that was dated and a page that was stamped. Three of the ten highest-ranked positions are not in this table: one returned a human-verification wall to both reading paths, one is a page whose text never reaches the document and whose browser reading timed out, and one was read in a browser but yielded no publication date, so it cannot be placed in a table sorted by date. The first two have no reading to contribute; the third has one and no row to put it in. None of the three is replaced with a substitute row. The span runs from the 1901 imprint to 31 March 2026, one hundred and twenty-five years; the dated web rows alone cover 1,447 days from 14 April 2022 to 31 March 2026, three years eleven months and seventeen days, and the oldest dated row inside the ten highest-ranked positions is 27 August 2024, which is 581 days before the newest. Where a quotation begins mid-sentence the break is marked.

What to ask of a page that will not tell you

A standard earring post length is decided by an ear and no face can change it. That is why every position above can print one and none of them can print the other number. Ask for the second figure instead, and ask for it in millimetres.

Ask four things, and ask for a number each time. From the piercing hole to the lowest point of the drop, how many millimetres? Does that figure include what hangs below the hook or sits behind the backing — one table in the set excludes both in its own column header, and most listings say nothing either way? For a hinged hoop, what is the inner diameter, and what is the lowest point, since those are two measurements and only one of them can be compared with a gap? And what does the answer become against your own gap rather than against the seller's word?

The last table inverts the same division so the answer is a length to order rather than a distance to estimate. Its columns are stops: the length that finishes level with the chin, and the lengths that finish half an inch above it or half an inch, an inch, or two inches below.

Each cell is the gap for that ratio at the working width, plus or minus the distance named in the column header, divided by 25.4 for the inch figure. The steps are chosen rather than measured and are marked as such: 12.7 mm is half an inch, 25.4 mm is one inch, and 50.8 mm is the corner-to-corner spread of the first table, from its smallest cell of 38.5 mm to its largest of 89.3 mm, so the last column is a drop that travels below the chin by exactly as far as the published band of faces differs from one corner of that table to the other. Because every cell is one gap plus a chosen distance, a reader working at another cheekbone width takes their row from the first table and adds the same four numbers to it. Two of these targets are worth naming in a shop: at ratio 1.35 the level stop is 61.2 mm, which is no product anyone sells, while the nearest thing to it is a 60 mm drop that finishes 1.2 mm short of the chin — the next size up, 70 mm, crosses by 8.8 mm. Two listings either side of a line you can compute.

Where this ladder is weakest

One assumption carries most of the risk in it. The gap is drawn from an earlobe sitting at the boundary between the middle and lower thirds of the face, and that is a declared position rather than a measured one: no position in the comparison puts the lobe at a level, and the pages that discuss lobes discuss their thickness. The cost is stated in the first table, not buried here — 6 mm up or down, added to every cell, which is more than the 4.5 mm step between two adjacent rows of that ladder, so a crossing can move a full row on the landmark alone. A study of ear height against facial landmarks would replace the assumption with a measurement and shift the crossing rows with it, and this page does not have one.

The width band is the second soft spot: 128.2 to 148.8 mm is the adult range this site publishes from one anthropometric dataset, and a band is not a measurement of you. Anyone outside it should use their own number, since the arithmetic is one multiplication and one division and the tables are built to be redone.

Third, the ladder ends where geometry ends. It reports where the lowest point of a drop is. A 40 mm line of stones and a 40 mm cluster terminate at the same number while covering quite different amounts of the face, and where an object's own width starts to matter against the width of a jaw is a separate question with separate numbers, not a row here.

Fourth, two ears are not one measurement, hair sits between a drop and a face, and a collar is not a chin. The gap is a reference line drawn on a bare face: it is where to start, not the last word on the day the thing is worn.