We show that, in a pioneering paper by Polnarev and Zembowicz, some conclus
ions concerning the characteristics of the Turok strings are generally not
correct. In addition we show that the probability of string collapse given
there is off by a large prefactor (similar to 10(3)).