A grafting sleeve has one job: hold two cut faces in exactly the same position for four weeks. It fails in two directions, and the two failures look identical at the count — a batch with poor take and no obvious cause.
Too big, and the join moves
New tissue across a graft union has no mechanical strength. It is dividing cells, replaced daily. If the sleeve allows even slight movement — a bench knocked, a tray lifted, a hose dragged past — the bridge being built is torn and rebuilt, torn and rebuilt. Some of those grafts still take, later and weaker than they should. Many do not.
The tell is a batch where take is mediocre across the board rather than concentrated in one area of the house. Environmental problems tend to have a geography. Fit problems do not.
Too small, and you press on the tissue that has to grow
The cambium is a layer perhaps a few cells thick, immediately under the surface. It is the only tissue that can make the join. Squeeze it and you slow or stop the division that the whole operation depends on.
A sleeve fitted too tightly also tends to mark the stem, and the mark is often read as “damage during handling” rather than as what it is.
So measure, and measure the right thing
Three rules:
Measure the thicker of the two. Scion and rootstock are rarely identical. The sleeve has to pass over both, so it is the larger diameter that sets the size. If you size to the scion and the rootstock is thicker, the sleeve will not seat.
Measure at the cut, not at the base. A rootstock cutting tapers. A measurement taken 40 mm down can be half a millimetre over what the sleeve will actually sit on — which is the difference between two sizes.
Measure a sample, not a stem. Take callipers to at least thirty stems drawn from across the batch, not thirty from the top of one crate. Write down the range, not the average. If the range spans more than about a millimetre, you need two sizes, and buying one is a decision to accept losses at both ends.
What tolerance means here
Every sleeve we supply is specified to a bore tolerance of ± 0.15 mm, held to batch. That is not a marketing number; it is the thing that decides whether a nominal 4.5 mm sleeve behaves like a 4.5 mm sleeve across a whole box.
It matters because your stems already vary. If the product varies by the same amount again, the two variations stack, and a proportion of the box is effectively the wrong size. A supplier who will not state a tolerance is asking you to absorb theirs.
Wall thickness is the other half
Bore decides whether it fits. Wall thickness decides how firmly it holds and how long it stays. A heavier wall grips harder and releases later; a lighter wall is gentler and gives up sooner. Our rose range runs 0.50 mm at the smaller bores and 0.70 mm at the larger, because a wider stem needs more material to exert the same pressure.
The vegetable sleeve is thinner again, at 0.40 mm, because a seedling stem a couple of millimetres across will not tolerate the pressure a rose cutting needs.
Why transparent, and why it is a specification
Our sleeves transmit 88 % of light or better. That is not cosmetic. It lets a supervisor look at a join without removing the sleeve, and the alternative — pulling sleeves to inspect — destroys a proportion of the batch being inspected. A specification that lets you check without touching pays for itself in the first week.
A working method
- Take callipers to thirty stems across the batch, at the point of cut.
- Record the range.
- Size to the larger of scion and rootstock.
- If the range exceeds roughly a millimetre, order two adjacent sizes and sort at the bench.
- Keep the box and lot code with the batch card, so a take figure can be traced back to a specific delivery.
That last step is the one everybody skips, and it is the only way you will ever know whether a change of supplier helped.