So, having been schooled by Ron about the change in the control for the CSV, a little investigation yields this:

http://volvo1800pictures.com/document/ES73%20Wiring%20digram/73%20Wiring.jpg

the diagram shows the thermal switch controlling the cold start valve directly - no relay.

As Ron points out, if you measure the voltage at the cold start valve with the starter motor engaged, 10.5 volts would be a very acceptable voltage because of the internal drop in the battery. The cold start valve is ground switched by the thermal timer device. I think the easy test would be to apply a ground at the thermal timer to force the cold start valve on during cranking by by-passing the thermal timer. If the cold start valve operates, then it would appear that the thermal timer device is faulty. Just for added confusion, the wire numbering (33 & 34) for the thermal timer connections is reversed between the 1973 wiring diagram for the 1800E and the 1973 Djet trouble shooting manual wiring diagram. I would normally refer to wire numbers; but, Volvo seems to have had some document control issues. According to the schematics, the thermal control device has two terminals. One has 2 wires and the other 1 wire. According to the wiring diagram, the connection to ground for testing would be the terminal with 1 wire. Note that if you ground the other wire you will create a dead short with very undesirable results so getting it right is important.

There was reference to the cold start valve not operating above 55C. This appears to be the setting for the earlier system controlled by the D jet controller. The Djet trouble shooting manual indicates that the operating point for the thermal timer switch is 35 C (pg 2-19). If you are doing work on the car in really hot weather, it might be possible that even though the engine is at ambient temperature, you might be above the operating point of the thermal timer. I am also pretty sure that the device is not exactly precision so it probably has a +/- 2 deg or more tolerance). As an observation, I wonder if the 35 C is a misprint because that does seem too low to provide zero starting enrichment.

Since my advice is guided purely by conflicting wiring diagrams, I defer to the testing advice of anybody who has actually done this.