Richard Aiken wrote:
 >> The purpose made neuromuscular blocking drugs
 >> (curare derivatives, aminosteroids, succinylcholine)
 >> can only be given by injection.
 >
 > How about injectors built into a cloud of nanobots?

You would need to be hit with a large number of nanobots,
which would have to be able to make the drugs with locally available
materials unless the bots carry enough to get you to the dose required.

Effective dose for neuromuscular blocking drugs is in the mg/kg body
mass range, so 10^(-3)g order of magnitude.

Bacteria - true nanomachines if the zinc micromotors in the linked
article (and ACS Nano paper) are nanomachines
  - weigh about a picogram or 10^(-12)g each.

That's one of the problems with nanotechnology as described in SF;
conservation of mass, thermodynamics and (chemical) kinetics are ignored.

Where does the waste (heat, chemical byproducts) from replication and
other reactions go?
Where does the material to make more copies or other substances come from?
How do nanomachines avoid attacks from the immune system or
environmental detritivores (bacteria and fungi)?

The real-world analog to nanomachine attack would be (viral) infection -
an incubation phase as machine numbers built up to the stage where they
could produce macroscopic effects, then an active phase of varying
intensity and duration.

Acute poisoning requires enough mass of toxin reaching target sites
quickly enough to swamp clearance mechanisms.

So the hard limit to onset speed is tens of seconds (circulation time).
Practically, an intravenous bolus injection is the only way to reliably
get close to this.

Meso- and microscale drones (mm length scale) are happening about now:
'Slaughterbots' for a 'day after tomorrow' extrapolation:
https://www.youtube.com/watch?v=TlO2gcs1YvM

Current military tech:
https://www.youtube.com/watch?v=5NGgHyfPGU0

Graham Donald wrote:
 > It need not be anything too fancy, reading the last few
 > posts I was reminded of the Dick Francis novel 'Trial Run',
 > in that novel the killers are using a powerful animal
 > tranquilizer (that has muscle relaxing properties.) mixed
 > with a skin penetrating agent.

It needs to be something quite fancy because the combination of
effects desired (i.e. voluntary muscle paralysis without affecting
other striated/skeletal muscle fibres and interfering with ventilation)
is infeasible, as Richard, Ken and I have pointed out.

The neurophysiology of terrestrial animals doesn't permit it -
our muscles and nervous systems are interlinked in such a way as to
forbid it.

A veterinary/toxicological aside:
Etorphine or carfentanil (two high-potency opioids used as
animal tranquillisers) don't need to be mixed with
a skin penetrating agent (dimethyl sulfoxide [DMSO] is beloved
of alties and mystery writers); they are very, very fat-soluble.

Darts as delivery systems are useful because they bypass the skin
and reduce the onset time (and perhaps the dose needed) compared
to the transcutaneous route.

Robert O'Connor