contracted to the individual. And, of course, if you are an individual member of the faculty, you can see the advantage in this arrangement.

Regarding your second question on who initiates strategic alliances, it can be either from the faculty up or management down. In the case of Air Products, it took place because one of our alumni was a vice president of research at Air Products and suggested that it would be a good idea to get back together at Imperial College. But in some cases, it's been a top-down decision by someone such as myself, saying that we should talk to Smithkline Beecham about analytical chemistry, which in our case disappeared into almost nothing inside the Chemistry Department. We wanted to build up again. It turned out that the department could see a national need for that because no one was being trained in analytical chemistry. Several pharmaceutical companies then came together and agreed with us. It wasn't sponsoring particular research projects, it was the existence of the discipline that was supported. So there has to be a synergy between what the companies want and what the college wants.

Janet Osteryoung, National Science Foundation: Have you run into any problems with some of your external partners bumping into each other inside the university?

William Wakeham: Yes. Our linkage with British Petroleum (BP), for example, which is in the area of catalysis, essentially supported about 30 people in one laboratory in a particular area. The type of catalyst being studied was of interest to more than one company. Imperial Chemical Industries (ICI) was also interested in that catalyst, but for a polymer that was of no interest to BP. So we tried very hard to get the companies to understand that they could work together. In the end it failed. ICI would not be involved because BP was already involved, even though the technologies coming out of the alliance would have been completely independent.