The cut happens on a Sunday because that's when you have the nerve. You have sterilized the blade, laid out sphagnum that's been soaking since morning, set the perlite aside, found the chopstick you use to tamp. The Philodendron on the shelf — a verrucosum from an Ecuadorian line, four leaves, the new one still half-furled and bronze — has thrown an aerial root the length of a finger. You have been watching it for weeks. Today is the day.
And then you stand there with the knife and realize you don't actually know which node to take. The top one, with the freshest growth? The middle, with the fat root? The basal node down by the substrate, hidden under leaf litter, smaller and more reluctant? They are not the same cutting. They will not behave the same in the box. One of them will root in three weeks. One of them will rot.
What a node actually is
A node is the structural intersection where the plant has, at some point in its life, made a leaf. On an aroid stem you can find it by the scar — a faint ring or a raised collar — and usually by two other features sitting nearby: an axillary bud (the meristem that will become the next growth point) and, in most climbers, a root primordium or an already-emerged aerial root. Internode is the stretch between. Internode tissue can callus; it cannot, on its own, build you a plant.
This is the first rule and it gets violated constantly. A length of bare stem with no node will sit in sphagnum for six months looking hopeful and then turn to mush. There is no bud to break. The cambium can callus over the wound, sometimes spectacularly, and people mistake that white tissue for a sign of life. It isn't. It's a scab.
The second rule is subtler: not every node is equally awake. The bud at a node has a dormancy gradient set by hormones produced higher up the stem — chiefly auxin, flowing down from the apex, suppressing lateral buds below it. This is apical dominance, and in a healthy unbroken stem it means the nodes nearest the growing tip are chemically muted. They look fine. They are not in a hurry.
A node is a decision, not a unit of stem. Cut the wrong one and you are propagating regret in clear plastic.
Why the prettiest node is the wrong one
Collectors gravitate to the node just below the newest leaf. It looks vigorous. The aerial root there is plump and green-tipped. The petiole scar is fresh. Of course you want that piece — it's the one that reads, visually, as the most alive part of the plant.
But that node is sitting directly under the apex's auxin shadow. Its axillary bud has been told, hormonally, to stay quiet. Cut it off the mother and yes, eventually the suppression lifts and the bud activates — but eventually can mean six weeks of waiting in a humid box where every day is another roll of the bacterial dice. Meanwhile the older node four down the stem, the one with a slightly shriveled aerial root and a barely visible bud, has been quietly accumulating reserves and ignoring the apex for months. Sever it and it breaks within ten days.
I learned this on an Anthurium warocqueanum that I cut at the obvious place — right under the top leaf, beautiful root, perfect petiole stub. It sat. The mother plant, with the basal nodes I'd left intact, threw two new growth points in a month. The cutting finally pushed a bud eleven weeks later and I had aged visibly in the meantime.
The geometry of the cut
Where you cut matters as much as which node you choose. The instinct is to slice tight against the node to keep the cutting compact. Resist it. Leave roughly a centimeter of internode tissue on either side of the node — above and below. That stub above protects the bud from desiccation and pathogen ingress; the stub below gives you a sacrificial zone if rot starts from the cut face, as it often does.
Use a blade, not scissors. Scissors crush. A single-edged razor or a scalpel makes a clean planar wound that callouses evenly. Sterilize between cuts with isopropyl or a quick pass through a flame — not because every plant in your collection is diseased, but because Erwinia and Pectobacterium are opportunists and a contaminated blade is how a single soft rot becomes a shelf-wide event.
Then let the cut dry. Twenty minutes on a paper towel, sometimes longer for thick-stemmed Philodendron gloriosum or Anthurium rhizomes that ooze. The callus that forms in that brief air-dry is a meaningful barrier. People who pot up wet wounds straight into wet sphagnum are running an experiment whose result is usually mush.
Which cuts rot
Rot in propagation is almost never random. It has causes, and once you've watched enough cuttings fail you can predict them. Soft, water-engorged tissue rots — so anything taken from a plant that was watered the day before is at elevated risk. Bruised tissue rots — so the node you handled three times deciding whether to cut it is compromised before it leaves the mother. Tissue without a clear callus rots. Tissue buried too deep in saturated sphagnum rots, because the bud sits in a pocket of anaerobic water.
Genus matters too. Philodendron stems are generally forgiving; their thick cortex tolerates a lot of operator error. Anthurium basal cuttings are middle ground — slow but steady if kept on the dry side of moist. Hoya stems with no nodes are a famous trap; people propagate the long leafy whip between nodes and wonder why nothing happens. Begonia rhizomes want surface contact, not burial, and rot fast if you press them into wet media.
The single best predictor of a failed cutting, in my notes across a few hundred attempts, is humidity without airflow. A sealed clear box at 95% RH and no air movement is a petri dish. Crack the lid daily. Better, run a small computer fan on low through a vent. Cuttings root in moving air at 70% humidity more reliably than they do in stagnant air at 90.
The basal node nobody wants
Down at the bottom of a mature climbing Philodendron — under the leaf litter, hidden by old cataphylls, often without a visible aerial root — there is usually a node that looks like nothing. Slightly woody. Faintly scarred. No green. This is, very often, the best cutting on the plant.
It has been awake for a long time. Its bud has been suppressed by apical dominance but the node itself has thickened and stored. Sever it with a generous internode buffer and put it on barely-damp sphagnum, and within two weeks you'll see the bud swell. Within four, a cataphyll. The first leaf out of a basal node is often larger than anything the cutting from above would have produced, because the reserves were already there.
The catch is that basal nodes are ugly and people don't trust them. They also require sacrificing a piece of the mother plant that doesn't visibly need to come off. There is a psychological tax to taking the unpretty cut. Pay it anyway.
Reading the mother plant first
Before you cut anything, look at the whole stem. Count nodes from the apex down. Note which ones have active aerial roots — meaning roots that are green-tipped and growing, not roots that emerged a year ago and have hardened off. Note which nodes have visible bud swelling: a small bump where the cataphyll meets the stem. Note which nodes are on the shaded side of the plant versus the lit side; shaded-side buds tend to be more dormant.
The cutting you want is the one that is already trying to do something. A swollen bud is the plant telling you, in advance, where it wants to branch. Helping it along by separating that node from the apex's suppression is a low-friction operation. Forcing a quiet node to wake up in a box is high-friction. Both work. Only one of them works on your timeline.
I keep a notebook with rough sketches of the stems of plants I propagate from often — a gloriosum rhizome map, the climbing stem of a melanochrysum, the basal cluster on a mature luxurians. It looks obsessive because it is. But it means that when the moment comes, on a Sunday afternoon with a sterilized blade and an hour of nerve, I am not standing over the plant guessing. I already know which node.
After the cut
The first week is the dangerous one. Put the cutting on damp — not wet — long-fiber sphagnum, or a 50/50 sphagnum-perlite mix if you're worried about moisture retention. The bud should sit at the surface, not buried. The cut faces should be in contact with media but not submerged. Bright indirect light, ambient warmth around 22–26°C, and that crucial trickle of air movement.
Resist the urge to check it every day by lifting and inspecting. Every handling is a chance to introduce contamination or break the new root hairs that are forming microscopically before you can see them. Look through the side of the container. If you must open it, do so to vent, not to fondle.
A successful cutting tells you it's working in a particular order: bud swells, cataphyll emerges, root tips appear, leaf unfurls. If you get bud activity without roots, you have a cutting living on reserves and you need to get roots going before those reserves run out — usually by lowering media moisture slightly and increasing airflow, which stresses the cutting in a productive direction. If you get roots without bud activity, you have a cutting that will sit indefinitely and you may need to wait, or take a second cut higher up that includes a more active bud. Propagation is, in the end, a slow conversation with a plant that answers in weeks. The skill is learning which questions to ask, and which nodes are capable of answering.